Persistent link: http://hdl.handle.net/2345/970

This work is posted on eScholarship@BC,Boston College University Libraries.

Boston College Electronic Thesis or Dissertation, 2008

Copyright is held by the author, with all rights reserved, unless otherwise noted.

Development of new methods forcatalytic enantioselective olefinmetathesis

Author: German Alexander Cortez

Boston College

The Graduate School of Arts and Sciences

Department of Chemistry

DEVELOPMENT OF NEW METHODS FOR CATALYTIC ENANTIOSELECTIVE OLEFIN METATHESIS

A dissertation

by

GERMAN ALEXANDER CORTEZ

submitted in partial fulfillment of the requirements

for the degree of

Doctor of Philosophy

Dec 2008

© copyright by German Alexander Cortez 2008

DEVELOPMENT OF NEW METHODS FOR CATALYTIC ENANTIOSELECTIVE OLEFIN METATHESIS

German Alexander Cortez

Thesis Adviser: Professor Amir H. Hoveyda

Abstract ♦ Chapter One. The synthesis of enantioenriched polycycles (in up to >98% ee) was accomplished through Mo-catalyzed asymmetric ring-opening/ring-closing metathesis (AROM/RCM) reactions of achiral norbornyl trienes. As a representative example, Mo complex ii catalyzes the formation of polycycle iii in 96% ee and 80% yield from triene i (shown below). These investigations were the first to show that an adamantylimido Mo alkylidene complex (ii) is capable of efficiently and selectively catalyzing enantioselective olefin metathesis transformations. These studies also provided insight into plausible catalytic reaction pathways for this class of metathesis transformations.

O O

H

HMo

N

PhMe

Me

OO

5 mol % ii

C6H6, 22 oC,3 h

i iii 96% ee, 80%

+

♦ Chapter Two. The preparation of chiral amines (up to 97% ee) was realized through the Mo-catalyzed asymmetric ring-closing metathesis (ARCM) reactions of achiral polyene precursors. These studies resulted in the development of methods to provide chiral cyclic amine products that contain readily removable protecting groups on nitrogen. Although generally considered sensitive to heteroatom functional groups, the Mo complexes in these studies were found to be tolerant of certain unprotected amine substrates (shown below). Mo complex v catalyzes the desymmetrization of secondary amine iv to provide cyclic amine vi in 71% ee and 95% yield.

5 mol % v

C6H6, 22 oC,24 h

Mo

N

MeMe

MeCl Cl

OO

NH Ph

iv

NH

Ph

vi 71% ee, 95%

+

♦ Chapter Three. Efforts resulting in the enantioselective synthesis of piperidines (up to >98% ee) through the Mo-catalyzed asymmetric ring-opening/cross-metathesis (AROM/CM) reactions of various meso azabicyclic substrates are discussed in this chapter. As an illustrative example, Mo complex ii catalyzes the desymmetrization of azabicycle vii with styrene to provide N-Me piperidine viii in 94% ee and 95% yield after 1 h. Enantioenriched carbamate-protected piperidines can also be prepared by this method, and moreover, these studies showed that a variety of aryl-olefin cross-partners can be employed for the desymmetrization reactions of azabicycles.

Mo

N

PhMe

Me

OO

5 mol % ii

C6H6, 22 oC,1 h

MeN

OTBS

NMe

OTBS

Ph

vii viii 94% ee, 95%10 equiv.

+

♦ Chapter Four. In this chapter we provide details of our investigations into the functionalization of piperidine products formed through asymmetric ring-opening/cross-metathesis (AROM/CM) reactions. Experiments were initially directed toward the synthesis of two biologically active piperidine alkaloids. These studies resulted in the discovery of a unique directed hydroalumination of styrenyl olefins. As shown below, this process was used to install oxygen functionality at a benzylic position by converting carbamate-protected piperidine ix to N-Me piperidine x in 45% yield. Additionally, studies were conducted to shed light into the mechanism of this novel transformation.

NCO2Et

OH2 equiv. LAH

THF, 65 oC, 2 h;O2

NMe

x 45%

OH OH

ix ♦ Chapter Five. In this chapter we describe studies undertaken to compare the ability of chiral Ru and Mo complexes to catalyze the formation of enantioenriched pyrans and piperidines. A representative example of the reactions studied is shown below. Ru complex xii catalyzed the AROM/CM reaction of N-Bn azabicycle xi with styrene to afford N-Bn piperidine xiii. In this example, Ru complex xii was found to be the optimal catalyst for the synthesis of xiii; a screen of available chiral Mo complexes resulted in the formation of xiii with <20% ee. In contrast, Mo complexes proved to be the optimal catalyst in the formation of N-Me piperidines (see above, synthesis of viii); Ru complexes provided viii in 30% ee and 20% yield. These investigations encompass the first direct comparison of chiral Ru and Mo complexes for asymmetric olefin metathesis.

2 mol % xii

neat, 22 oC,24 h

BnN

OTBS

NBn

OTBS

Ph

xi xiii >98% ee, 80%<20% ee with Mo complexes

Ru

N

O

NCl

Oi-Pr

Ph

Me

MeMe

20 equiv.

+

i

TABLE OF CONTENTS

CHAPTER 1: MOLYBDENUM-CATALYZED ENANTIOSELECTIVE SYNTHESIS OF POLYCYCLES THROUGH ASYMMETRIC RING-OPENING/RING-CLOSING METATHESIS REACTIONS

1.1 Introduction .................................................................................................1 1.2 Mo-catalyzed Enantioselective Synthesis of Bicycles and Tricycles .......11 1.3 An Example of a Dynamic Asymmetric Process ......................................16 1.4 Limitation of Mo-catalyzed AROM/RCM Method ..................................18 1.5 Conclusions ...............................................................................................19 1.6 Experimental Section ................................................................................20

CHAPTER 2: MO-CATALYZED ENANTIOSELECTIVE SYNTHESIS OF N-CONTAINING HETEROCYCLES THROUGH ASYMMETRIC RING-CLOSING METATHESIS

2.1 Introduction ...............................................................................................43 2.2 Asymmetric Ring-Closing Metathesis of Protected Amines ....................47 2.3 Asymmetric Ring-Closing Metathesis of Unprotected

Secondary Amines ....................................................................................49 2.4 Reactivity of Ru Complexes With Amine-Containing Substrates ............53 2.5 Conclusions ...............................................................................................53 2.6 Experimental Section ................................................................................54

CHAPTER 3: MO-CATALYZED ENANTIOSELECTIVE SYNTHESIS OF PIPERIDINES THROUGH ASYMMETRIC RING-OPENING/CROSS-METATHESIS

3.1 Introduction ...............................................................................................84 3.2 AROM/CM approach to N-Me Piperidines ..............................................87 3.3 Mo-catalyzed AROCM of Carbamate-Protected

Azabicycles ...............................................................................................96 3.4 Mo-catalyzed AROM/CM of Bn-protected Azabicycles .........................99 3.5 Mo-catalyzed AROM/CM of Azabicycles with Non-Aryl

Olefin Cross-Partners ..............................................................................100 3.6 Conclusions .............................................................................................102 3.7 Experimental Section ..............................................................................103

ii

CHAPTER 4: FUNCTIONALIZATION OF PIPERIDINE PRODUCTS

4.1 Introduction .............................................................................................189 4.2 Initial Approach to Piperidine Alkaloids ................................................190 4.3 Discovery of a Unique Directed Hydroalumination ...............................193 4.4 Alternative Approach to Piperidine Alkaloids .......................................198 4.5 Miscellaneous Functionalization Reactions of Piperidines ...................200 4.6 Conclusions .............................................................................................201 4.7 Experimental Section ..............................................................................202

CHAPTER 5: COMPARISON OF MOLYBDENUM AND RUTHENIUM COMPLEXES FOR CATALYTIC ASYMMETRIC RING-OPENING CROSS-METATHESIS REACTIONS

5.1 Introduction .............................................................................................231 5.2 Second Generation Ru Catalyst for the Enantioselective

Synthesis of Pyrans .................................................................................233 5.3 Mo Complexes for the Enantioselective

Synthesis of Pyrans .................................................................................235 5.4 Ru Complexes for the Enantioselective Synthesis of

Carbamate-protected Piperidines ............................................................238 5.5 Ru Complexes for the Enantioselective Synthesis of N-alkyl

Piperidines ..........................................................................................240 5.6 Conclusions .............................................................................................244 5.7 Experimental Section ..............................................................................245

Chapter 1, Page 1

Chapter One

MOLYBDENUM-CATALYZED ENANTIOSELECTIVE SYNTHESIS OF POLYCYCLES THROUGH ASYMMETRIC RING-OPENING/RING-CLOSING

METATHESIS REACTIONS

1.1 Introduction

My research endeavors during my doctoral candidacy were largely focused on the development of new methods involving catalytic enantioselective olefin metathesis. When I began my graduate career efforts were focused on the application of chiral Mo-based complexes to enantioselective olefin metathesis. The advent of chiral Ru-based catalysts for olefin metathesis, which occurred as I was starting my graduate work, allowed me the opportunity to (in the latter part of my graduate career) study these complexes in addition to Mo-based complexes. With this in mind, I decided to begin this introduction by providing a timeline of Mo-catalyzed enantioselective olefin metathesis reactions that were developed by the collaborative work between the Schrock and Hoveyda groups. It is hoped that this timeline will enlighten the reader as to the progression of asymmetric olefin metathesis, from its inception to the time when my research began. In doing so, the hope is also that the reader will acquire an understanding of attributes innate to chiral Mo-based complexes employed for olefin metathesis. This discussion should also facilitate an understanding of concepts to be presented in future chapters relating to comparative studies between chiral Mo- and Ru-based complexes for olefin metathesis.

1.1a Enantioselective Synthesis of Cyclopentenes and Dihydrofurans. Our first collaborative report of a Mo-catalyzed enantioselective olefin metathesis process was published in 1998.1 Initial studies involved the kinetic resolution of 1,6-dienes through ring-closing metathesis reactions. The metathesis of 1,6-dienes was chosen partly because it was well documented that Mo complexes were efficient at performing such ring closures; products and byproducts of the reactions were known and could be readily identified.2 Furthermore, the olefin substitution patterns of such 1,6-dienes were easily

(1) "Catalytic Enantioselective Ring-Closing Metathesis by a Chiral Biphen-Mo Complex," Alexander, J. B.; La, D. S.; Cefalo, D. R.; Hoveyda, A. H.; Schrock, R. R. J. Am. Chem. Soc. 1998, 120, 4041−4042. (2) “Synthesis of Cycloalkenes Via Alkylidene-mediated Olefin Metathesis and Carbonyl Olefination,” Fu, G. C.; Grubbs, R. H. J. Am. Chem. Soc. 1993, 115, 3800−3801.

Chapter 1, Page 2

modified, and allowed us to predict the site of initiation by Mo complexes. The biphenol-containing catalyst 1.2 was previously prepared by the Schrock group and chosen for these studies due to its efficiency in controlling the stereochemistry of ring-opening metathesis polymerization (ROMP) reactions of norbornenes.3 A representative example of the outcome from these studies is illustrated in eq. 1.1. Exposure of 1,6-diene 1.1 to 5 mol% Mo complex 1.2 delivers cyclopentene 1.3 in 93% ee (krel = 58)4 and 43% yield after a reaction time of only 10 min.5 Enantioenriched 1,6-diene 1.1 is isolated in >98% ee and 19% yield.

Mo

N

PhMe

Mei-Pr i-Pr

OO

Me

MeOTES

C6H6, 22 oC,10 min

5 mol % 1.2 Me

OTES

1.1 1.3 93% ee, 43%

Me

MeOTES

1.1 >98% ee, 19%

(1.1)+

A mechanistic model accounting for formation of cyclopentene 1.3 is shown in

Figure 1.1. Mo alkylidene 1.2 likely reacts at the terminal olefin of the starting 1,6-diene and adopts a pseudo-chair conformation prior to the subsequent metathesis that furnishes 1.3. It is clear from this mechanistic model that the major enantiomer that forms has a lower activation energy barrier (∆G††) in its transition state than the minor enantiomer. If reaction with the slower-reacting (minor) enantiomer were to occur, severe steric repulsion would be present between the pendant silyl ether and the catalyst complex.

Mo

Ni-Pr i-Pr

OO

OTES

Figure 1.1 Proposed TransitionState Model for Kinetic Resolution

(3) “Ring Opening Metathesis Polymerization with Binaphtholate or Biphenolate Complexes of Molybdenum,” Totland, K. M.; Boyd, T. J.; Lavoie, G. G.; Davis, W. M.; Schrock, R. R. Macromolecules 1996, 29, 6114−6125. (4) The value for krel is calculated by the equation reported by Kagan, see: Kagan, H. B.; Fiaud, J. C. Top. Stereochem. 1988, 18, 249−330. (5) After ten minutes the reaction has proceeds to 81% conversion to product. The facility of this ring-closure attests to the high reactivity of these high-oxidation state metal complexes.

Chapter 1, Page 3

Although highly effective in the kinetic resolution to form cyclopentene products, this method was not amenable to the formation of cyclohexene products; kinetic resolution of 1,7-dienes proceeds to krel values <11.6 These limitations of this method made it imperative that more selective chiral complexes be developed. Additionally, since these were the first ventures into asymmetric olefin metathesis it remained unknown what other classes of asymmetric transformations could be efficiently and selectively catalyzed by Mo complex 1.2.

After studying kinetic resolution reactions, we focused on to investigate the desymmetrization reactions of meso 1,6-dienes. In general, catalytic desymmetrization processes are more atom-economical than kinetic resolutions. In a desymmetrization reaction the product can be isolated in up to 100% yield, in contrast to a maximum yield of 50% in a kinetic resolution. One major goal of these studies was to determine whether chiral Mo complexes could differentiate between enantiotopic olefins in a meso compound. To this end, we studied the Mo-catalyzed desymmetrization of meso 1,6-trienes such as the representative example shown in eq. 1.2.7 Treatment of triene 1.4 with 2 mol % Mo complex 1.5 in the absence of solvent leads to the formation of dihydrofuran 1.6 in >98% ee and 93% yield within 5 min.

Mo

N

PhMe

MeMe Me

OO

MeO

neat, 22 oC,5 min

2 mol % 1.5

1.4 1.6 >98% ee, 93%

(1.2)MeO

Me

Me

Me Me Me

Several facets of this chemical transformation deserve discussion. First, the

reaction involves minimal waste production. The metathesis can be performed without

(6) Values of krel >50 are generally required if products of high enantiomeric purity are to be obtained in useful yield, see: “Practical Considerations in Kinetic Resolution Reactions,” Keith, J. M.; Larrow, J. F.; Jacobsen, E. N. Adv. Synth. Catal. 2001, 343, 5−26. (7) "Mo-Catalyzed Asymmetric Synthesis of Dihydrofurans. Catalytic Kinetic Resolution and Enantioselective Desymmetrization through Ring-Closing Metathesis," La, D. S.; Alexander, J. B.; Cefalo, D. R.; Graf, D. D.; Hoveyda, A. H.; Schrock, R. R. J. Am. Chem. Soc. 1998, 120, 9720−9721.

Chapter 1, Page 4

solvent and the only byproduct from the reaction is propene, a volatile and easily removable gas. Second, the optimal catalyst for the reaction (1.5) is a chiral Mo complex that was newly developed; previously discussed Mo complex 1.2 provides 1.6 in >98% ee, however, in only 32% conversion after 9 h. These results show that subtle differences in catalyst structure, in this case changing the imido ligand from 2,5-i-Pr2 in catalyst 1.2 to 2,5-Me2 in catalyst 1.5, lead to significant reactivity differences. In short, these observations highlighted the need to have available a family of various structurally dissimilar, yet highly reactive, chiral Mo complexes. Most importantly, these studies provided a new method for the enantioselective synthesis of heterocyclic compounds. Heterocyclic compounds are invaluable intermediates for therapeutics8 and this method allows for the preparation of enantiopure 2-substituted dihydrofuran, intermediates not easily accessible by other known methods.

1.1b Enantioselective Synthesis of Six-membered Ring Compounds through Catalytic Metathesis. The next set of studies involved desymmetrization reaction protocols for the formation of six-membered ring heterocycles, exemplified in Scheme 1.1 (see below).9 As previously mentioned, difficulties were encountered in trying to perform desymmetrization transformations of 1,7-dienes. Mo complex 1.8, a new catalyst construct containing a bidentate binaphtholate ligand, was prepared in response to the potential need for catalyst architecture different from that of biphenol-based Mo complexes 1.2 and 1.5. Utilizing this catalyst, exposure of siloxy-containing triene 1.7 to 2 mol % of Mo complex 1.8 in the absence of solvent at 60 oC produces silyl ether 1.9 in >98% ee and 98% yield after 4 h. Under identical reaction conditions, dihydropyran 1.11 is isolated in >98% ee and 86% yield after 3 h. As in the synthesis of chiral dihydrofurans, Mo-catalyzed reactions were performed in the absence of solvent. Notably, reactions proceed to high levels of enantioselectivity even at elevated temperature. These experiments revealed that Mo-catalyst 1.8 was the best catalyst for the synthesis of chiral six-membered ring intermediates; Mo complexes 1.2 and 1.5 generate the desired products with lower efficiency and in lower selectivity than complex 1.8. Although complex 1.8 was found to be the optimal catalyst for the desymmetrization of 1,7-dienes, the complex was less effective for the desymmetrization of 1,6-dienes such as formation of 1.6 (see eq. 1.2 above), again highlighting the advantage of sustaining a

(8) Pozharskii, A. F.; Soldatenkov, A. T.; Katritzky, A. R. in Heterocycles in Life and Society, Wiley−VCH, Weinheim, Germany, 1997; Edition 1. (9) "Chiral Mo-Binol Complexes: Activity, Synthesis, and Structure. Efficient Enantioselective Six-Membered Ring Synthesis through Catalytic Metathesis," Zhu, S. S.; Cefalo, D. R.; La, D. S.; Jamieson, J. Y.; Davis, W. M.; Hoveyda, A. H.; Schrock, R. R. J. Am. Chem. Soc. 1999, 121, 8251−8259.

Chapter 1, Page 5

family of chiral Mo complexes. After achieving success in various asymmetric ring-closing metathesis transformations, we moved on to study a completely different class of olefin metathesis reactions, namely, the ring-opening/cross-metathesis reaction of meso norbornenes.

MeO

neat, 60 oC,4 h

1.7 1.9 >98% ee, 98%

Me

Me2Si

O

Ar

OOMo

MePh

MeN

i-Pri-Pr

Ar

2 mol % 1.8Ar = 2,4,6,-i-Pr3(C6H2) O

Me2Si

Me

Me

MeO

n-pentane, 60 oC,3 h

1.10 1.11 >98% ee, 86%

Me2 mol % 1.8 O

Me

Me

Scheme 1.1

1.1c Asymmetric Ring-Opening Metatheses Processes. A beautiful feature−if not the most−of the olefin metathesis reaction is the number of possible permutations of metatheses that can be performed under the right reaction conditions. A highly strained norbornene, for example, in the presence of Mo-catalysts typically undergoes polymerization reactions; polymerization reactions of norbornenes are well-studied and have received much attention from chemists in recent years.10 It is generally accepted that the primary driving force for ring-opening reactions of norbornenes is release of ring strain. In the presence of an appropriate olefin donor, however, polymerization reactions of norbornenes can be prevented and instead lead to the formation of cyclopentanes through ring-opening/cross-metathesis reactions.11 To explore this possibility with chiral

(10) For a review of ring-opening metathesis polymerization reactions, see: Black, G.; Maher, D.; Risse, W. in Handbook of Metathesis (Ed: R. H. Grubbs), Wiley−VCH, Weinheim, Germany, 2003; vol. 3, pp. 2−71. (11) For a review of catalytic ring-opening olefin metathesis reactions, see: Schrader, T. O.; Snapper, M. L. in Handbook of Metathesis (Ed: R. H. Grubbs), Wiley−VCH, Weinheim, Germany, 2003; vol. 2, pp. 205−237.

Chapter 1, Page 6

Mo complexes we studied the desymmetrization of meso norbornyl ethers; a representative example of these studies is shown in eq. 1.3. In the presence of two equivalents of styrene and 5 mol % of Mo complex 1.2, norbornyl ether 1.12 is desymmetrized to form enantioenriched cyclopentane 1.13 in >98% ee and 85% yield after 1 h. From an initial inspection it would appear that this is a facile reaction, however, it should be noted that ring-opening/cross-metathesis reactions are often difficult to control and predict. Ring-opening/cross-metathesis reactions are bimolecular reactions that suffer from formation of byproducts that would not be present in other metathesis processes, for example ring-closing metathesis reactions.

2 equiv. styrene,C6H6, 22 oC,

1 h

OTMS

1.12 1.13 >98% ee, 85%

(1.3)

OTMS

Mo

N

PhMe

Mei-Pr i-Pr

OO

5 mol % 1.2

Difficulties encountered in developing ring-opening cross-metathesis reactions can be better appreciated through a discussion of potential and observed mechanistic pathways; the Mo intermediates and potential byproducts involved in the desymmetrization of norbornene 1.12 are shown in Scheme 1.2 (see below). Initially, Mo benzylidene i is formed through reaction of the Mo pre-catalyst neopentylidene (1.2 in eq. 1.3) with styrene. Studies from our group have shown that Mo benzylidenes are slow to react with another equivalent styrene, as evident from minimal stillbene formation in the absence of another olefin cross partner (norbornene in this example). Ring-opening of norbornene 1.12 with Mo benzylidene i thus leads to formation of Mo alklylidene ii. This newly generated alkylidene can react with another equivalent of norbornene or styrene. Reaction with norbornene would form Mo alkylidene iii, which represents an intermediate that would be involved in norbornene polymerization. Alternatively, reaction of alkylidene ii with styrene furnishes cyclopentane 1.13, the observed product of the reaction, and regenerates Mo benzylidene i. At this point, if benzylidene i reacts at the terminal olefin of the newly generated cyclopentene 1.13, then bis-cross metathesis product iv would be formed along with Mo methylidene v. Studies

Chapter 1, Page 7

from the Shrock group have shown that Mo methylidene intermediates are highly reactive (in relation to benzylidene i) and have been implicated in catalyst decomposition pathways.12 In addition to catalyst decomposition, Mo methylidene intermediates can be involved in non-productive metathesis pathways; if Mo methylidene v were to react with norbornene 1.12, the result would be formation of Mo alkylidene vi. Intermediate vi in turn could react with another equivalent of norbornene 1.12 and provide a Mo alkylidene similar to that of iii (this intermediate would have a terminal olefin in place of a styrenyl olefin at its terminus). Alternatively, alkylidene vi could react with styrene leading to meso cyclopentane viii and regenerate Mo benzylidene i. It should be noted that in the ring-opening/cross-metathesis being discussed (eq. 1.3, see above) byproducts iv and viii are not observed; formation of non-chiral cyclopentanes such as iv and viii are sensitive to the identity of the starting norbornene and meso intermediates of this type have been observed in other ring-opening cross-metathesis reactions. In concluding this discussion on the mechanism of ring-opening cross-metathesis reactions, it should also be noted that not all potential pathways were shown. Pathways leading to other byproducts, such as oligomers or products arising from an initial ring-opening of a Mo pre-catalyst neopentylidene have also been observed in other reactions.

(12) “Reduction of Molybdenum Imido-alkylidene Complexes in the Presence of Olefins to Give Molybdenum (IV) Complexes,” Robbins, J.; Bazan, G. C.; Murdzek, J. S.; O’Regan, M. B.; Schrock, R. R. Organometallics 1991, 10, 2902−2907.

Chapter 1, Page 8

OTMS

LnMo+

OTMS

MoLn

OTMS

MoLn

TMSO

OTMS

LnMo

OTMS OTMS

LnMo

LnMo

OTMS

LnMo

1.12i ii

iii

i

1.13

i

iv v

vi

OTMS

1.12

OTMS

viii

Scheme 1.2

1.13

LnMo

v

LnMo

i

+

+

1.1d Tandem Ring-Opening/Ring-Closing Metathesis Reactions. The next class of asymmetric reactions developed was the ring-opening/ring-closing metathesis of cyclobutenes for the formation of enantioenriched dihydrofurans.13 Hitherto, enantioenriched dihydrofurans were prepared through asymmetric ring-closing

(13) "Tandem Catalytic Asymmetric Ring-Opening Metathesis/Ring-Closing Metathesis," Weatherhead, G. S.; Ford, J. G.; Alexanian, E. J.; Schrock, R. R.; Hoveyda, A. H. J. Am. Chem. Soc. 2000, 122, 1828−1829.

Chapter 1, Page 9

metathesis reactions. A different approach to this class of chiral compounds is through a skeletal rearrangement exemplified in eq. 1.4. Treatment of cyclobutene 1.14a with 5 mol % Mo complex 1.2 delivers dihydrofuran 1.15a in 92% ee and 69% yield after 1 h. When the metathesis of a related cyclobutene, terminal olefin containing 1.14b, was studied the desired product is formed in >98% ee, however, in only 10% yield. The major product from the reaction of 1.14b is meso double ring-closed product 1.15c. This difference in product distribution highlights the sensitivity of Mo complexes to the olefin substitution pattern of metathesis substrates.

C6H6, 22 oC,1 h

1.14a R = Me

1.14b R = H

1.15a 92% ee, 69%

1.15a >98% ee, 10%

(1.4)O O

R RO

R

O

R

Mo

N

PhMe

Mei-Pr i-Pr

OO

5 mol % 1.2 OO

1.15c

As previously mentioned, the release of strain in a molecule can serve to energetically favor a metathesis reaction. Along these lines, after the desymmetrization of cyclobutenes we moved on to study the ring-opening/ring-closing rearrangement of cyclopentenes to deliver cyclohexenes; a representative example of this class of asymmetric transformations is shown in eq. 1.5 (see below). Exposure of cyclopentene 1.16 to 5 mol % of Mo complex 1.17 delivers enantioenriched cyclohexene 1.18 in 78% ee and >98% conversion (consumption of substrate as monitored by 1H NMR analysis) after 24 h.14 In light of the moderate selectivity observed in this reaction (78% ee) we examined various additives in attempts to improve the enantioselectivity of this reaction. Previous studies with Mo complexes, in particular studies with 1.17, revealed that Lewis basic additives such as THF were capable of altering the equilibrium between syn and anti benzylidene isomers. It was hoped that the altering this equilibrium could be beneficial to the selectivity of the reaction. When 2 equiv. of THF are used as an additive (14) "Enhancement of Enantioselectivity by THF in Asymmetric Mo-Catalyzed Olefin Metathesis. Catalytic Enantioselective Synthesis of Cyclic Tertiary Ethers and Spirocycles," Teng, X.; Cefalo, D. R.; Schrock, R. R.; Hoveyda, A. H. J. Am. Chem. Soc. 2002, 124, 10779−10784.

Chapter 1, Page 10

in this reaction 1.18 is isolated in 85% ee and >98% conversion. When the optimal amounts of THF additive (10 equiv.) are used for this reaction the product 1.18 is isolated in 92% ee, >98% conversion, and 93% isolated yield after 24 h. As we have seen, our studies up to this point have shown that a wide variety of factors must be taken into consideration when developing a catalytic enantioselective olefin metathesis method; this in effect teaches and challenges us as chemists to “not leave a stone unturned” when trying to solve problems in organic chemistry.

C6H6, 4 oC,24 h

1.16

no THF 78% ee, >98% conv.2 equiv. THF 85% ee, >98% conv10 equiv. THF 92% ee, >98% conv. 93%

(1.5)

O

Ar

OOMo

MePh

MeN

ClCl

Ar

5 mol % 1.17Ar = 2,4,6,-i-Pr3(C6H2)MOMO

MOMO

1.18

1.1e Enantioselective Synthesis of Medium-size Rings. Our next investigations dealt with the preparation of enantioenriched seven-membered ring heterocycles, intermediates previously inaccessible through Mo-catalyzed olefin metathesis.15 One of the difficulties in performing ring-closing metathesis reactions of larger ring sizes is the tendency of metathesis catalysts to undergo cross-metathesis reactions, delivering substrate homodimers, instead of the desired ring-closed products. A way of circumventing this problem is by relying on the reversibility of cross-metathesis reactions, since cyclic olefins of low strain energy are less likely to undergo a reversible ring-opening metathesis reaction. In particular, the problem of troublesome ring-closures of macrocycles has been solved in this manner. 16 It is more difficult to get around the

(15) "Enantioselective Synthesis of Medium-Ring Heterocycles, Tertiary Ethers, and Tertiary Alcohols by Mo-Catalyzed Ring-Closing Metathesis," Kiely, A. F.; Jernelius, J. A.; Schrock, R. R.; Hoveyda, A. H. J. Am. Chem. Soc. 2002, 124, 2868−2869. (16) “Applications of Zr-Catalyzed Carbomagnesation and Mo-Catalyzed Macrocyclic Ring Closing Metathesis in Asymmetric Synthesis. Enantioselective Total Synthesis of Sch 38516 (Fluvirucin B1),” Xu,

Chapter 1, Page 11

problem of cross-metathesis when medium size rings are the desired metathesis products. The approach we took to solve this problem, which has also been used by others for related ring-closures,17 was to incorporate functional groups in the substrate molecule that would induce a Thorpe-Ingold effect,18 and thus facilitate a ring-closing metathesis to occur. Toward this end, treatment of triene 1.19 with 5 mol % of Mo complex 1.20 in benzene delivers seven-membered heterocycle 1.21 in 87% ee and 95% yield after 15 min. Two additional aspects of this reaction merit mention: (1) the reaction was performed with 1 g of substrate, demonstrating that reactions with Mo complexes can be scaled to multi-millimol quantities (3.2 mmol in this example). Second, the product can be subjected to proteo-desilylation reaction conditions to provide an enantioenriched tertiary alcohol (not shown) that would otherwise be difficult to prepare.

O

1.19 (1.0 g) 1.21 87% ee, 95%

(1.6)SiMe2

Mo

N

MeMe

MeCl Cl

OO

C6H6, 22 oC,15 min

5 mol % 1.20Ph

Me2Si

OPh

1.2 Mo-catalyzed Enantioselective Synthesis of Bicycles and Tricycles

And this is where my story, of the contributions I made to catalytic enantioselective olefin metathesis reactions, begins. It was a cold winter day, on January 1st, 2002, when I entered the labs of Professor Amir H. Hoveyda to begin my graduate career. Gabriel Weatherhead, an utmost thoughtful chemist, greeted me and it was only a matter of

Z.; Johannes, C. W.; Houri, A. F.; La, D. S.; Cogan, D. A.; Hofilena, G. E.; Hoveyda, A. H. J. Am. Chem. Soc. 1997, 119, 10302−10316. (17) “A Concise Total Synthesis of Dactylol via Ring Closing Metathesis,” Fürstner, A.; Langemann, K. J. Org. Chem. 1996, 61, 8746−8749. (18) The Thorpe-Ingold effect, or gem-dimethyl effect or angle compression, is phenomena observed in organic chemistry where increasing the size of two substituents on a tetrahedral arrangement around a center leads to enhanced reactions between parts of the other two substituents. For the original publication on this effect, see: (a) Beesley, R. M.; Ingold, C. K.; Thorpe, J. F. J. Chem. Soc. 1915, 107, 1080−1106. For additional discussion on this topic, see: (b) “Conformational Analysis. IX. The Gem-Dimethyl Effect,” Allinger, N. L.; Zalkow, V. J. Org. Chem. 1960, 25, 701−704.

Chapter 1, Page 12

moments before our objectives were clearly defined and we began our search for chemicals with which to commence our studies; I was to work with Gabe on the development of a new class of tandem ring-opening/ring-closing metathesis reactions.

1.2a Desymmetrization Reactions of Meso-Norbornenes. Our attention was focused on the possibility of performing asymmetric ring-opening/ring-closing metathesis (AROMRCM) reactions of meso-norbornenes. Previous work had showed that norbornenes were compatible substrates for ring-opening/cross-metathesis reactions (vide supra). But what if instead of having an intermolecular reaction between a norbornene and olefin-cross partner, we had an achiral norbornene with a tethered olefin that could react in an intramolecular fashion, would it be possible to perform tandem ring-opening/ring-closing metathesis reactions? The substrates chosen to answer these questions are shown in Scheme 1.3. Our goal was to prepare substrate types, exemplified by substrate 1.22 and 1.24, which could lead to the formation of structurally complex bicycles (1.23) or tricycles (1.25).

OMo-Catalyst

O

OR

H

R

H

O

R

H

Scheme 1.3

1.22 R = H or alkyl 1.23 R = H or alkyl

Mo-Catalyst

1.24 R = H or alkyl 1.25 R = H or alkyl

R

1.2b AROM/RCM of Norbornyl Trienes. As illustrated in Scheme 1.4 (see

below), our studies began by investigating the desymmetrization of norbornyl triene 1.24 to provide tricycle 1.25.19 Initial screening conditions employed 5 mol % of chiral Mo-complexes (see Chart 1 below for catalysts screened)20 and reactions were quenched after 3 h. As previously discussed, one potential drawback to olefin metathesis transformations is for nonproductive reaction pathways to occur. A byproduct observed during our screen is achiral seven-membered ring 1.26, arising from the ring-closing (19) The synthesis of 1.28 and all other substrates is described in the experimental section of this chapter. (20) For convenience, from this point forward chiral Mo catalysts are numbered alphabetically.

Chapter 1, Page 13

metathesis between the two terminal olefins in substrate 1.24. As illustrated by the graph in Scheme 1.4, Mo-catalysts demonstrate varying levels of selectivity and reactivity for the desymmetrization of 1.24. Two notable examples of such differences are exemplified by catalysts I and J. Whereas I exclusively catalyzes formation of achiral 1.26, the major product formed in the reaction with J is the desired tricycle 1.25. Under optimized reaction conditions with Mo complex J, tricycle 1.25 is isolated in 96% ee and 80% yield.21 At the time complex J was the newest addition to our family of chiral Mo complexes and these studies resulted in the first example of complex J catalyzing a highly selective olefin metathesis reaction. Complex J is structurally very different from previous Mo complexes in that J possesses an adamantylimido ligand, in contrast to earlier complexes that contain arylimido ligands.

O

H

OH5 mol % chiralMo-Catalyst

1.24 1.25

C6H6, 22 oC,3 h

Scheme 1.4

1.26

O+

H

OH

96% ee, 80% 1.25and 20% 1.26

with J

0

20

40

60

80

100

% c

onv.

/ % e

e

A B C D E F G H I J

1.251.26

% ee 1.25

Catalysts Screened

(21) “An Enantiomerically Pure Adamantylimido Molybdenum Alkylidene Complex. An Effective New Catalyst for Enantioselective Olefin Metathesis,” Tsang, W. C. P.; Jernelius, J. A.; Cortez, G. A.; Weatherhead, G. S.; Schrock, R. R.; Hoveyda, A. H. J. Am. Chem. Soc. 2003, 125, 2591−2596.

Chapter 1, Page 14

Mo

N

R'Me

MeR R

OO

A R = i-Pr, R' = PhB R = Cl, R' = MeC R = Me, R' = Ph

O

Ar

OOMo

MeR'

MeN

RR

Ar

Mo

N

PhMe

Me

OO

J

D R = i-Pr, R' = PhE R = Cl, R' = MeF R = Me, R' = PhAr = 2,4,6,-i-Pr3(C6H2)

O

Ar

OOMo

MePh

MeN

RR

Ar

H R = Cl, Ar = CHPh2I Ar = CHPh2

Chart 1. Catalysts Screened for AROMRCM of Norbornyl Trienes

O

Ar

OOMo

MePh

MeN

CF3

Ar

G Ar = 2,4,6,-i-Pr3(C6H2)

The next substrate studied was norbornyl triene 1.27, a diastereomer of 1.24. As shown in eq. 1.7, after screening available chiral Mo-catalysts, the optimal catalyst for the desymmetrization of 1.27 was again adamantylimido Mo complex J. Exposure of 1.27 to 5 mol % of J provides tricycle 1.28 in 78% ee and 90% isolated yield after 3 h. One notable difference in the formation of 1.28 (vs. 1.25) is that the desired product is formed in lower enantioselectivity (78% ee for 1.28 vs. 96% ee for 1.25). Another notable difference is that the achiral byproduct 1.29, which would form through a ring-closing metathesis of the terminal olefins, is not observed. These results highlight the difficulties in predicting the outcome of any given metathesis reaction.

OO

H

H5 mol % J

1.27 1.28 78% ee, 90%

CH2Cl2, 22 oC,3 h

(1.7)

1.29not observed

O+

Mo

N

PhMe

Me

OO

Chapter 1, Page 15

1.2c Mechanistic Model for AROM/RCM of Norbornyl Trienes. A mechanistic rationale for the product distribution difference between the desymmetrization of 1.24 and 1.27 is shown in Scheme 1.5.

MoLnO

1.24b

O

MoLn

H

OH

1.25

MoLn

O

1.24a

LnMo

O

LnMo

1.27a

O

MoLn

O

H

H

1.28

or

+

Scheme 1.5

1.26

O

i

iii

ii

MoLn+

ii

MoLn

ii

+

Studies from several groups, including ours, have described the chelating effect of

proximal heteroatoms in metathesis reactions.22 Presumably, the active Mo catalyst first reacts with the allylic ether olefin in 1.27 resulting in intermediate 1.27a where an O→Mo chelation is present. The alkylidene of 1.27a then reacts with the norbornyl olefin to provide intermediate i, furnishing one of the fused six membered rings present in 1.28. Finally, ring-closure with the pendant terminal olefin provides tricycle 1.28 along with Mo-methylidene ii. Based on the preceding supposition that a Mo-alkylidene

(22) (a) “Reaction of Neopentylidene Complexes of the Type M(CH-t-Bu)(N-2,6-C6H3-i-Pr2)(OR)2 (M = W, Mo) with Methyl Acrylate and N,N-dimethylacrylamide To Give Metallacyclobutane Complexes,” Feldman, J.; Murdzek, J. S.; Davis, W. M.; Schrock, R. R. Organometallics 1989, 8, 2260−2265. (b) Dolman, S. New Chiral Molybdenum Metathesis Catalysts; Application To The Enantioselective Preparation of Cyclic Amines. Ph. D. Dissertation, Massachusetts Institute of Technology, Cambridge, MA, June 2004.

Chapter 1, Page 16

preferentially reacts at an allylic ether olefin over an alkyl olefin, in the reaction of 1.24 with Mo complex J, two alkylidene intermediates are possible, 1.24a and 1.24b. Intermediate 1.24b would follow a reaction pathway similar to that of 1.27a, by first providing iii and ultimately delivering 1.25 along with Mo-methylidene ii. In intermediate 1.24a, the spatial distance separating the norbornyl olefin and the reactive alkylidene make it unlikely that the two can undergo metallacyclobutane formation. It is at this point that reaction of the alkylidene in 1.24a with the pendant terminal olefin likely occurs resulting in the formation of seven-membered ring product 1.26. This mechanistic proposal would require the alkylidene in 1.24a to undergo a different metathesis reaction, such as reaction with another molecule of 1.24 to arrive at 1.24b. Based on this mechanistic analysis, we decided to prepare a substrate with a disubstituted allylic ether olefin distal to the norbornyl olefin. We reasoned that steric differentiation should favor reaction of the Mo alkylidene at the terminal olefin (similar to formation of 1.24b) and thus prevent seven-membered ring product formation.

1.3 An Example of a Dynamic Asymmetric Process

1.3a Observation of a Bicyclic Ring-Opening/Ring-Closing Product. To this end, we prepared and studied the desymmetrization of norbornyl triene 1.30 (eq. 1.8). Upon screening of available chiral catalysts, we found Mo complex B to be the optimal catalyst in the desymmetrization of 1.30: Treatment of 1.30 with 5 mol % Mo catalyst B furnishes tricycle 1.31 in >98% ee and 75% yield after 3 h. Mo complex J was also selective in the formation of 1.31 (>98% ee); however, the product is isolated in only 60% yield.

O

H

OH5 mol % B

1.30 1.31 >98% ee, 75%

C6H6, 22 oC,3 h

(1.8)+

MeMe

H

OH

Me

1.32 15%

Mo

N

PhMe

MeCl Cl

OO

As predicted, increasing the substitution at the allylic ether olefin disfavors formation

of a seven-membered ring product; however, a different byproduct, bicycle 1.32, is

Chapter 1, Page 17

isolated in 15% yield (complex J provides 1.32 in 40% yield). This was not surprising since the metathesis to form cyclic tri- and tetrasubstituted olefins is generally sluggish; in comparison to reactions for the formation of cyclic disubstituted olefins. These results, at least in part, substantiate our mechanistic proposal for the AROM/RCM of this class of norbornyl trienes.

1.3b Determination of a Catalytic Dynamic Asymmetric Process. Given the high selectivity in the formation of 1.31 and the formation of 1.32, we speculated as to whether or not this reaction was an example of a matched vs. miss-matched enantioselective process (Scheme 1.6). If, based on our mechanistic model, the initial Mo alkylidene reacts at the terminal olefin, followed by ring-opening of the norbornene, an intermediate such as 1.33 should be formed along the reaction coordinate. Is it possible that the ring-closure of 1.33 is faster than the ring-closure of ent 1.33, the enantiomer of 1.33? If the ring-closure of ent 1.33 is slow and instead leads to formation of bicycle 1.32, then the enantiopurity of 1.31 should be high, as we have observed. To address this question we would have to determine the enantiomeric excess of 1.32. Ring-closure of 1.32 to tricycle 1.31 would provide the answer to this question. Indeed, when 1.32 is treated with 5 mol % of achiral Mo complex 1.34, ent 1.31 is isolated in –76% ee and 80% yield.

H

OH

LnMoMe

1.33

H

OH

LnMoMe

ent 1.33

H

OH

1.31

Me

H

OH

ent 1.31

Me

f ast slow

Scheme 1.6

5 mol % 1.34

C6H6, 22 oC,3 h

H

OH

Me

1.32

OOMo

N

i-Pri-Pr

MeCF3CF3

F3CF3C Me

ent 1.31 −76% ee, 80%

PhMe

Me

H

OH

Me

Chapter 1, Page 18

1.4 Limitation of Mo-catalyzed AROM/RCM Method

1.4a Kinetic Resolution Approach to Natural Product Synthesis. To demonstrate the practicality of our newly developed method, we undertook the synthesis of the sesquiterpene natural product africanol, isolated from the marine invertebrate Lemnalia africana.23 During our studies, Gabe and I took two different approaches toward africanol. My studies involved the kinetic resolution of norbornene 1.35 to access bicycle 1.36 (Scheme 1.7). We postulated that the kinetic product of this ring-opening/ring-closing rearrangement should be 1.36, which would posses an exocyclic 1,1-disubstituted olefin. Then, under the proper reaction conditions a metathesis rearrangement should allow for the formation of the thermodynamic product 1.37, which would possess an endocyclic trisubstituted olefin. Functional group manipulations would then provide (−)-africanol. Unfortunately, under various reaction conditions we were never able to perform the desired kinetic resolution of norbornene 1.35; in all cases the starting material was recovered in >98% mass recovery.

Mo CatalystTBSO

Me

Me Me

H

OTBS

H

Scheme 1.7

1.35

(-)-africanol

OH

H

MeMe

Me

Me

OTBS MeMeMe

Me

MeMe

1.36

1.37

MetathesisRearrangement

1.4b Desymmetrization Approach to Natural Product Synthesis. The alternative approach to africanol proved to be more fruitful (Scheme 1.8, see below). This approach involved the desymmetrization of norbornene 1.38. To this end, Gabe found that treatment of norbornene 1.38 with 3 mol % of Mo complex 1.17 in pentane delivers the

(23) “Chemical studies of marine invertebrates. VIII africanol, an unusual sesquiterpene from Lemnalia africana (Coelenterata, Octocorallia, Alcyonacea),” Tursch, B.; Braekman, J. C.; Daloze, D.; Fritz, P.; Kelecom, A.; Karlsson, R.; Losmon, D. Tetrahedron Lett. 1974, 15, 747−750.

Chapter 1, Page 19

desired bicycle 1.39 in 87% ee and 97% yield after 6 h.24 After several functional group manipulations, Gabe was able to arrive at an intermediate that was previously reported en route toward a total synthesis of africanol,25 constituting an enantioselective formal synthesis.

TBSO

(+)-africanol

Me Me

Scheme 1.8

OH

H

MeMe

Me

Me

H

OTBS MeMe

1.39 87% ee, 97%

pentane, 22 oC, 6 h

O

Ar

OOMo

MePh

MeN

i-Pri-Pr

Ar

3 mol % 1.17Ar = 2,4,6-i-Pr3(C6H2)

1.38

1.5 Conclusions

These studies resulted in the development of a variety of Mo-catalyzed tandem asymmetric ring-opening/ring-closing metathesis reactions. Previously, norbornenes had only been substrates in asymmetric ring-opening/cross-metathesis reactions to deliver cyclopentanes. With these investigations, we were able to show that olefin metathesis can take a single class of substrates, meso norbornenes, and form a variety of enantioenriched products (bicycles and tricycles), demonstrating the versatility of olefin metathesis for organic synthesis. These studies also demonstrate for the first time that a Mo alkylimido alkylidene complex (J, see above) is capable of efficiently and selectively catalyzing olefin metathesis reactions. In addition to developing new methods for olefin metathesis, these studies provided insight into the reaction mechanism of such

(24) “Mo-catalyzed asymmetric olefin metathesis in target-oriented synthesis: Enantioselective synthesis of (+)-africanol,” Weatherhead, G. S.; Cortez, G. A.; Schrock, R. R.; Hoveyda, A. H. Proc. Nat. Acad. Sci. 2004, 101, 5805−5809. (25) Fan, W.; White, J. B. J. Org. Chem. 1993, 58, 3557−3562.

Chapter 1, Page 20

transformations. Arguably, the most important aspect of this work was that it demonstrated that catalytic olefin metathesis is applicable to complex natural product synthesis.

1.6 Experimental Section

Materials and Reagents.26 All reactions were conducted in oven- (135 oC) or flame-dried glassware under an inert atmosphere of dry N2 unless otherwise stated. Infrared (IR) spectra were recorded on a Nicolet 210 spectrophotometer, νmax in cm-1. Bands are characterized as broad (br), strong (s), medium (m) or weak (w). 1H NMR spectra were recorded on a Varian Gemini 2000 (400 MHz) spectrometer. Chemical shifts are reported in ppm from tetramethylsilane with the solvent resonance as the internal reference (CDCl3: δ 7.26, C6D6: δ 7.16). Data are reported as follows: chemical shift, multiplicity (s = singlet, d = doublet, t = triplet, q = quartet, br = broad, m = multiplet or combinations thereof), coupling constants (Hz), and integration. 13C NMR spectra were recorded on a Varian Gemini 2000 (100 MHz) spectrometer with complete proton decoupling. Chemical shifts are reported in ppm from tetramethylsilane with the solvent resonance as the internal reference (CDCl3: δ 77.16, C6D6: δ 128.10). Enantiomer ratios were determined by chiral HPLC (Chiral Technologies Chiralpak OD column, Chiralpak AD, Chiralcel OJ, and Chiralcel OB-H (4.6 mm x 250 mm)). High-resolution mass spectrometry was performed at the University of Illinois Mass Spectrometry Laboratories (Urbana-Champaign, IL). Elemental microanalyses were performed by Robertson Microlit Laboratories (Madison, N. J.). Optical rotation values were recorded on a Rudolph Research Analytical Autopol IV polarimeter.

Solvents were purged with argon and then purified under a positive pressure of dry argon by a modified Innovative Technologies purification system: CH2Cl2 was passed through activated alumina columns; benzene was passed successively through activated Cu and alumina columns. Tetrahydrofuran was purified by distillation from sodium benzophenone ketal immediately prior to use. All reagents were purchased from Aldrich Chemical Company, Lancaster Synthesis, or Strem Chemicals, Inc., and purified by appropriate methods prior to use. Lithium aluminum hydride was purchased in powder form and was stored in a N2-filled glovebox. Mo complex J was prepared according to published procedures.16 Mo complexes were handled under an inert atmosphere in a N2-filled glovebox. Substrates employed in Mo-catalyzed reactions were dried by repeated (three times) azeotropic distillation of water with benzene under high vacuum unless otherwise stated. Ru complexes 1a-2b were prepared according to published (26) This section on materials and reagents applies to subsequent chapters as well.

Chapter 1, Page 21

procedures.27 All reactions were allowed to stir with a magnetic stir bar and performed at 22 oC, unless otherwise stated. All filtrations involved gravity filtration; gravity filtrations were conducted with Whatman® filter papers. All reagent solvents were purchased from Doe and Ingalls.

O 2.4 equiv. t-BuLi

1.2 equiv. Br

Et2O+

1.40

OHHO

1.41 1.42

(1.9)

Synthesis of acohols 1.41 and 1.42. Alcohols 1.41 and 1.42 were prepared from norbornenone28 as shown in eq. 1.8. A two-neck 500-mL round-bottom flask, fitted with an addition funnel, was charged with 4-bromobutene (3.08 mL, 30.4 mmol, 1.2 equiv.), Et2O (250 mL), and the reaction vessel was cooled to −78 °C in a dry ice/acetone bath. To this solution was added t-butyllithium (33.22 mL, 60.80 mmol, 1.83 M solution in hexane) dropwise over a period of 2 h. At this time, the reaction vessel was warmed to 0 °C and allowed to stir for 1 h, after which time it was cooled to −78 °C and allowed to stir a further 4 h. At this moment, norbornenone 1.40 (2.70 g, 25.0 mmol) was added to this mixture by cannula, after which time the cooling bath was removed and the reaction mixture was allowed to stir for 12 h. The reaction was quenched by the addition of H2O (50 mL), the layers were separated, and the aqueous layer was washed with Et2O (3 x 50 mL). The combined organic layers were dried (MgSO4), filtered, and the volatiles removed in vacuo. The resulting dark brown residue was purified by AgNO3-impregnated silica gel chromatography (gradient elution, 1:6 to 1:3 Et2O:hexanes) to give 1.41 as colorless oil (0.918 g, 5.59 mmol, 22%) and 1.42 as colorless oil (0.890 g, 5.40 mmol, 20%); the identity of 1.41 was confirmed by nOe analysis. Norbornene 1.41: 1H NMR (400 MHz, CDCl3): δ 6.12 (dd, J = 2.0, 2.0 Hz, 2H), 5.86 (dddd, J = 17.2, 10.4, 6.8, 6.8 Hz, 1H), 5.03 (ddd, J = 17.2, 1.6, 1.6 Hz, 1H), 4.93 (dddd, J = 10.4, 1.2, 1.2, 1.2 Hz, 1H), 2.61−2.57 (m, 2H), 2.20−2.13 (m, 2H), 1.80−1.75 (m, 2H), 1.62−1.58 (m, 2H),

(27) (a) “A Recyclable Chiral Ru Catalyst for Enantioselective Olefin Metathesis. Efficient Catalytic Asymmetric Ring-Opening/Cross Metathesis in Air,” Van Veldhuizen, J. J.; Garber, S. B.; Kingsbury, J. S.; Hoveyda, A. H. J. Am. Chem. Soc. 2002, 124, 4954−4955. (b) “A Readily Available Chiral Ag-Based N-Heterocyclic Carbene Complex for Use in Efficient and Highly Enantioselective Ru-Catalyzed Olefin Metathesis and Cu-Catalyzed Allylic Alkylation Reactions,” Van Veldhuizen, J. J.; Campbell, J. E.; Giudici, R. E.; Hoveyda, A. H. J. Am. Chem. Soc. 2005, 127, 6877−6882. (28) Norbornenone was prepared in five steps through known chemical procedures. “The Chemistry of 7-Substituted Norbornenes,” Gassman, P. G.; Pape, P. G. Tetrahedron Lett. 1963, 4, 9−12.

Chapter 1, Page 22

0.96 (dd, J = 10.4 3.2 Hz, 2H). 13C NMR (100 MHz, CDCl3): δ 135.4, 132.0, 129.2, 63.0, 47.5, 29.4, 25.9, 23.8. Norbornene 1.42: 1H NMR (400 MHz, CDCl3): δ 5.98 (dd, J = 2.4, 2.4 Hz, 2H), 5.81 (dddd, J = 16.8, 10.0, 6.4, 6.4 Hz, 1H), 5.01 (dddd, J = 16.8, 1.6, 1.6, 1.6 Hz, 1H), 4.92 (dddd, J = 10.0, 2.0, 1.2, 1.2 Hz, 1H), 2.43 (ddd, J = 4.0, 2.0, 2.0 Hz, 2H), 2.12−2.06 (m, 2H), 1.98−1.92 (m, 2H), 1.80−1.75 (m, 2H), 0.97 (dd, J = 11.2, 3.6 Hz, 2H).

Allylation of alcohols 1.41 and 1.42 provided metathesis substrates 1.24, 1.27, and 1.30. Norbornene 1.24: IR (Neat): 2967 (m), 2942 (m). 1H NMR (400 MHz, CDCl3): δ 6.00 (dd, J = 2.0, 2.0 Hz, 2H), 5.90−5.79 (m, 2H), 5.19 (dddd, J = 17.2, 2.0, 1.6, 1.6 Hz, 1H), 5.06 (dddd, J = 10.4, 2.0, 1.2, 1.2 Hz, 1H), 5.01 (dddd, J = 17.2, 3.6, 1.6, 1.6 Hz,

1H), 4.92 (dddd, J = 10.0, 2.0, 1.2, 1.2 Hz, 1H), 3.72 (ddd, J = 2.8, 1.6, 1.6 Hz, 2H), 2.66 (ddd, J = 3.6, 2.0, 2.0 Hz, 2H), 2.14−2.07 (m, 2H), 1.70−1.67 (m, 2H), 2.14−2.07 (m, 2H), 1.57 (ddd, J = 8.4, 4.8, 4.8 Hz, 2H), 0.98 (ddd, J = 13.6, 2.8, 2.8 Hz, 2H). 13C NMR (100 MHz, CDCl3): δ 139.5, 135.6, 134.8, 115.8, 113.9, 96.5, 64.0, 46.8, 29.8, 25.8, 23.4.

Elem. Calcd for C14H20O: C, 82.30; H, 9.87. Found C, 82.11; H, 9.99.

O

1.24

Norbornene 1.27: 1H NMR (400 MHz, C6D6): δ 5.81−5.71 (m, 4H), 5.25 (dddd, J = 17.2, 4.0, 2.0, 2.0 Hz, 1H), 5.00−4.91 (m, 2H), 4.90 (dddd, J = 10.0, 2.0, 1.2, 1.2 Hz, 1H), 3.53 (ddd, J = 3.6, 1.6, 1.6 Hz, 2H) 2.41 (ddd, J = 3.6, 3.6, 2.0 Hz, 2H), 2.06−1.98

(m, 4H), 1.73−1.69 (m, 2H), 0.86 (ddd, J = 9.6, 3.2, 3.2 Hz, 2H). 13C NMR (100 MHz, CDCl3): δ 139.5, 135.7, 132.6, 116.2, 114.2, 97.5, 67.1, 46.7, 28.4, 28.3, 23.0. Elem. Calcd for C14H20O: C, 82.30; H, 9.87. Found C, 82.17, H, 10.00.

O

1.27

Norbornene 1.30: 1H NMR (400 MHz, CDCl3): δ 5.98 (dd, J = 2.0, 2.0 Hz, 2H), 5.80 (dddd, J = 16.8, 10.0, 6.4, 6.4 Hz, 1H), 5.10−4.95 (m, 2H), 4.89 (dddd, J = 10.4, 2.4, 1.6, 1.6 Hz, 1H), 4.85−4.83 (m, 1H), 3.67 (br s, 2H), 2.62 (dd, J = 3.6, 2.0 Hz, 2H), 2.05−1.98 (m, 2H), 1.96−1.91 (m, 2H), 1.90−1.76 (m, 5H), 0.90

(ddd, J = 10.8, 3.6, 3.6 Hz, 2H). 13C NMR (100 MHz, CDCl3): δ 143.1, 139.5, 134.8, 113.9, 111.1, 96.3, 66.7, 46.8, 29.9, 25.8, 23.4, 20.0. Elem. Calcd for C15H22O: C, 82.52; H, 10.16. Found C, 82.49, H, 10.27.

O

1.30

Me

Representative procedure for tandem Mo-catalyzed AROM/RCM; synthesis of tricycle 1.25 and meso 1.26. In a N2-filled glovebox, a 4-mL vial was charged with triene 1.24 (25.0 mg, 0.122 mmol) and C6H6 (1.20 mL). To this solution was added J (5.0 mg, 0.0061 mmol, 0.05 equiv) and the vial was fitted with a teflon lined cap. The

Chapter 1, Page 23

reaction mixture was allowed to stir at 22 °C for 3 h. At this time, the vial was removed from the glovebox, and the volatiles were removed in vacuo. The resulting brown residue was purified by silica gel chromatography (40:1 pentane:Et2O) to give 1.25 as colorless oil (17.7 mg, 0.0976 mmol, 80%) and 1.26 as colorless oil (4.3 mg, 0.024 mmol, 20%). Tricycle 1.25. IR (Neat): 3024 (m), 2955 (m), 1465 (w), 1440 (w); 1H NMR (400 MHz, C6D6): δ 5.78−5.70 (m, 3H), 5.30 (dddd, J = 10.0, 5.6, 2.4, 2.4 Hz, 1H), 4.18 (dddd, J = 17.6, 5.6, 2.8, 2.8 Hz, 1H), 4.01 (dddd, J = 17.6, 4.4, 2.4, 2.4 Hz, 1H), 2.64−2.54 (m, 1H), 2.24−2.06 (m, 2H), 1.91−1.77 (m, 3H), 1.55−1.47 (m, 1H), 1.26−1.00 (m, 3H). 13C NMR (100 MHz, CDCl3): δ 129.1, 126.3, 126.1, 125.9, 64.1, 45.2, 43.0, 29.9, 29.1, 23.5, 21.7, 19.1.

Elem. Calcd C12H16O: C, 81.77; H, 9.15. Found C, 81.84; H, 9.26. Meso 1.26. IR (Neat): 2961 (m), 2936 (m). 1H NMR (400 MHz, C6D6): δ 5.82 (s, 2H), 5.70−5.60 (m, 1H), 5.42−5.38 (m, 1H), 3.91 (s, 2H), 2.49 (s, 2H), 2.16 (d, J = 3.6 Hz, 2 H), 2.02 (s, 2H), 1.40−1.20 (m, 2H), 0.98 (d, J = 3.2 Hz, 1H), 0.85 (dd, J = 2.8, 2.8 Hz, 1H). 13C NMR (100 MHz, C6D6): δ 135.3, 131.9, 129.2, 97.6, 63.0, 47.4, 29.4, 25.9, 23.7). The optical purity of 1.25 was determined by chiral GC analysis in comparison with authentic racemic material, shown below: CDGTA column, 105 oC, 20 psi, [α]D

20 −75.91 (c = 1.2, CHCl3) for a sample of 96% ee.

H

OH

1.25

1.26

O

Tricycle 1.28. 1H NMR (400 MHz, C6D6): δ 5.82−5.68 (m, 3H), 5.43−5.39 (m, 1H), 3.96−3.84 (m, 2H), 2.25−2.13 (m, 2H), 2.12−1.97 (m, 2H), 1.96−1.74 (m, 1H), 1.73−1.58 (m, 2H), 1.28−1.14 (m, 2H), 0.92−0.84 (m, 1H). 13C NMR (100 MHz, C6D6): δ 127.2, 126.3, 124.5, 77.1, 65.9, 60.6, 48.1, 41.2, 30.7, 26.5, 26.3, 24.3. The optical purity of 1.28 was determined by chiral GC analysis in comparison with authentic racemic

material, shown below: CDGTA column, 70 oC for 180 min and then 140 oC 10 min thereafter (increasing at the rate of 70 oC/min), 20 psi, for a sample of 78% ee.

O

H

H

1.28

Chapter 1, Page 24

Tricycle 1.31. 1H NMR (400 MHz, C6D6): δ 5.80−5.76 (m, 2H), 5.44 (dddd, J = 5.6, 4.0, 2.0, 2.0 Hz, 1H), 4.08 (dddd, J = 17.2, 3.2, 2.4, 1.2 Hz, 1H), 3.97−3.90 (m, 1H), 2.64−2.59 (m, 1H), 2.34−2.21 (m, 2H), 1.98−1.72 (m, 3H), 1.58−1.1.50 (m, 1H), 1.32−1.03 (m, 6H). 13C NMR (100 MHz, CDCl3): δ 132.5, 129.1, 126.2, 121.0, 67.4, 45,3, 42.7, 29.5, 23.6, 21.6, 19.1, 18.2. Elem. Calcd C13H18O: C, 82.06; H, 9.53. Found C, 82.0; H, tical purity of 1.31 was determined by chiral GC analysis in comparison

with authentic racemic material, shown below: CDGTA column, 105 oC for, 20 psi, for a sample of >98% ee.

H

OH

1.31

Me

9.66. The op

Bicycle 1.32. 1H NMR (400 MHz, C6D6): δ 5.86−5.77 (m, 1H), 5.72−5.68 (m, 1H), 5.65−5.61 (m, 1H), 5.18−5.17 (m, 1H), 5.05−5.03 (m, 1H), 5.01−4.99 (m, 1H), 4.89−4.87 (m, 1H), 3.82 (dd, J = 12.4, 12.4 Hz, 2H), 2.75 (dd, J = 8.8, 8.8 Hz, 1H), 2.50−2.42 (m, 1H), 2.30−2.21 (m, 1H), 2.10−1.53 (m, 3H), 1.53 (s, 3H), 1.46−1.22 (m, 4H). The optical purity of ent 1.31, isolated from the reaction shown in Scheme

1.6, was determined by chiral GC analysis in comparison with authentic racemic material, shown below: CDGTA column, 105 oC, 20 psi, for a sample of −76 ee.

H

OH

Me

1.32

Chapter 1, Page 25

Chapter 1, Page 26

Chapter 1, Page 27

Chapter 1, Page 28

Chapter 1, Page 29

Chapter 1, Page 30

Chapter 1, Page 31

Chapter 1, Page 32

Chapter 1, Page 33

Chapter 1, Page 34

Chapter 1, Page 35

Chapter 1, Page 36

Chapter 1, Page 37

Chapter 1, Page 38

Chapter 1, Page 39

Chapter 1, Page 40

Chapter 1, Page 41

Chapter 1, Page 42

Chapter 2, Page 43

Chapter Two

MO-CATALYZED ENANTIOSELECTIVE SYNTHESIS OF N-CONTAINING HETEROCYCLES THROUGH ASYMMETRIC RING-CLOSING METATHESIS

2.1 Introduction

After being involved with a successful effort aimed at developing methods for the catalytic enantioselective synthesis of polycycles, I focused my attention to the development of methods for the catalytic enantioselective synthesis of heterocyclic amines.29

2.1a First Example of a Ring-Closing Metathesis of a Tertiary Amine Containing Substrate. Grubbs and Fu were the first to publish on the Mo-catalyzed ring-closing metathesis of tertiary amines and a representative example of this work is shown in eq. 2.1 (see below).30 Treatment of tertiary amine 2.1 with 4 mol % achiral Mo complex 2.2 affords heterocycle 2.3 in 73% yield after 1 h. Importantly, Mo catalysts typically outperform Ru-based complexes in olefin metathesis reactions involving substrates with tertiary amine functional groups.31

(29) (a) "Efficient Catalytic Enantioselective Synthesis of Unsaturated Amines: Preparation of Small- and Medium-Ring Cyclic Amines through Mo-Catalyzed Asymmetric Ring-Closing Metathesis in the Absence of Solvent," Dolman, S. J.; Sattely, E. S.; Hoveyda, A. H.; Schrock, R. R. J. Am. Chem. Soc. 2002, 124, 6991−6997. (b) "Enantioselective Synthesis of Cyclic Secondary Amines through Mo-Catalyzed Asymmetric Ring-Closing Metathesis (ARCM)," Dolman, S. J.; Schrock, R. R.; Hoveyda, A. H. Org. Lett. 2003, 5, 4899−4902. (30) “Synthesis of Nitrogen Heterocycles via Catalytic Ring-Closing Metathesis of Dienes,” Fu, G. C.; Grubbs, R. H. J. Am. Chem. Soc. 1992, 114, 7324−7325. (31) (a) “Catalytic Ring-Closing Metathesis of Functionalized Dienes by a Ruthenium Carbene Complex,” Fu, G. C.; Nguyen, S. T.; Grubbs, R. H. J. Am. Chem. Soc. 1993, 115, 9856−9857. (b) “Novel entry to the Ergot alkaloids via ring closing metathesis,” Lee, K. L.; Goh, J. B.; Martin S. F. Tetrahedron Lett. 2001, 42, 1635−1638. (c) “An Efficient Approach to Aspidosperma Alkaloids via [4 + 2] Cycloadditions of Aminosiloxydienes: Stereocontrolled Total Synthesis of (±)-Tabersonine. Gram-Scale Catalytic Asymmetric Syntheses of (+)-Tabersonine and (+)-16-Methoxytabersonine. Asymmetric Syntheses of (+)-Aspidospermidine and (−)-Quebrachamine,” Kozmin, S. A.; Iwama, T.; Huang, Y.; Rawal, V. H. J. Am. Chem. Soc. 2002, 124, 4628−4641.

Chapter 2, Page 44

2.1

NBn

4 mol % 2.2

C6H6, 20 oC,1 h

OOMo

MePh

MeN

i-Pri-Pr

MeCF3

CF3

F3C

F3C Me

Me

BnN

Me

2.3 73%

(2.1)

Mo complexes for olefin metathesis exist at the highest possible oxidation state,

that of VI, at the metal center. This makes Mo complexes highly Lewis acidic and sensitive to Lewis basic functional groups such as carbonyls and alcohols. Ru complexes, conversely, are in a low oxidation state at the metal center, that of +II, and are tolerant of Lewis basic functionalities. A priori, it would be expected for Mo complexes to be sensitive to tertiary amines, which are a class of Lewis basic compounds; Ru complexes, in turn would be expected to be tolerant of such functional groups. Tertiary amines, however, serve as compatible ligands for and tend to inhibit the metathesis activity of Ru complexes. Nonbonding electrons on tertiary amines have a high electron density coefficient (high lying Highest Occupied Molecular Orbital-HOMO) and have a greater affinity toward the less electropositive Ru complexes (low lying Lowest Unoccupied Molecular Orbital-LUMO) than toward the more electropositive Mo complexes (high lying LUMO).32 This discussion of functional group compatibility also serves to explain why compounds of high lying HOMOs (such as phosphines and N-heterocyclic carbenes) are good ligands for Ru complexes, whereas compounds of low lying HOMOs (such as alkoxides) are good ligands for Mo complexes. Based on the supposition that Mo complexes would be tolerant of tertiary amine functional groups, we initiated studies into the enantioselective synthesis of amine containing heterocycles.

2.1b Mo-catalyzed Asymmetric Ring-Closing Metathesis of Aniline Substrates. In collaboration with the Schrock group, we previously published on the first Mo-catalyzed enantioselective synthesis of amines. Initial efforts focused on the desymmetrization of aniline containing trienes and a representative example is shown in eq. 2.2 (see below).33

(32) For a discussion of HOMO-LUMO interactions, see: Fleming, I. in Frontier Orbitals and Organic Chemical Reactions, John Wile & Sons, Ltd. Chichester, England, 1976; Edition 1 pp 34−84. (33) "Efficient Catalytic Enantioselective Synthesis of Unsaturated Amines: Preparation of Small- and Medium-Ring Cyclic Amines through Mo-Catalyzed Asymmetric Ring-Closing Metathesis in the Absence

Chapter 2, Page 45

Exposure of aniline 2.4 to 4 mol % of Mo complex 2.5 delivers azocine 2.6 in 97% ee and 98% yield after 7 h.

2.4 2.6 97% ee, 98%

Mo

N

MeMe

MeCl Cl

OO

neat, 22 oC,7 h

4 mol % 2.5

Me NPh

Me

NPh

Me

Me(2.2)

One notable feature of this reaction is that the metathesis reaction can be

performed in the absence of solvent; exclusion of solvent from metathesis reactions minimizes waste production. Additionally, an eight-membered ring heterocycle can be prepared in high optical purity; in ring-closing metathesis reactions it is typically more difficult to form rings of greater size than six atom units (see Ch1, 1.1 Introduction).

2.1c Mo-catalyzed Asymmetric Ring-Closing Metathesis of Secondary Aniline Substrates. The next set of studies in Mo-catalyzed enantioselective syntheses of N-containing heterocycles involved the desymmetrization of anilines such as 2.7, illustrated in eq. 2.3 (see below).34 These studies were aimed at determining whether a desymmetrization reaction could be performed to deliver an enantioenriched heterocycle that contains a fully substituted carbon stereocenter. Thus, treatment of aniline 2.7 with 2 mol % of Mo complex 2.8 furnishes azepine 2.9 in 93% ee and 75% yield after 3 h. An attractive feature of this class of transformations is that the enantioenriched products would otherwise be difficult to access by other known catalytic enantioselective methods. Another aspect of this reaction worthy of mention is that the isolated product, aniline 2.9, contains a free N-H moiety. Mo complexes, due to the high reactivity at the alkylidene carbon, are typically not tolerant of functional groups containing Brønsted acids such as alcohols and carboxylic acids. In the presence of Brønsted acid functional groups, Mo

of Solvent," Dolman, S. J.; Sattely, E. S.; Hoveyda, A. H.; Schrock, R. R. J. Am. Chem. Soc. 2002, 124, 6991−6997. (34) "Enantioselective Synthesis of Cyclic Secondary Amines through Mo-Catalyzed Asymmetric Ring-Closing Metathesis (ARCM)," Dolman, S. J.; Hoveyda, A. H.; Schrock, R. R. Org. Lett. 2003, 5, 4899−4902.

Chapter 2, Page 46

complexes decompose through protonation at the alkylidene carbon. These experiments demonstrated that chiral Mo complexes are more functional group tolerant than previously believed.

C6H6, 55 oC,3 h

2.7 2.9 93% ee, 75%

O

Ar

OOMo

MePh

MeN

i-Pri-Pr

Ar

2 mol % 2.8Ar = 2,4,6,-Me3(C6H2)NH

Me

MeNH

Me

MeEt

Et (2.3)

At this point in my graduate career I became involved in a group effort that included Elizabeth Sattely (then a graduate student) and David Moebius (then an undergraduate student), aimed at further developing Mo-catalyzed asymmetric ring-closing metathesis reactions amines. We sought to study the desymmetrization reaction of non-aryl amines. Previous work was able to furnish a variety of arylamine chiral compounds, however, the phenyl group is not typically considered a protecting group for nitrogen.35 Within this context, we prepared and investigated the desymmetrization of protected amine trienes typified by 2.10 that would deliver enantioenriched cyclic amines such as 2.11 (eq. 2.4).

NP

R

Me

Me

NP

Me

Me

RMo Catalyst

(2.4)

2.10P = protecting group

R = alkyl2.11

(35) Conversion of arylamines to the derived unprotected amines remains an inefficient procedure and few examples have been reported in the literature; for a representative example, see: “Bradsher Cycloaddition of 4-Alkoxyisoquinolinium Salts as a Route to a Fully Functionalized B Ring of the Angucycline Antibiotics,” Nicolas, T. E.; Franck, R. W. J. Org. Chem. 1995, 60, 6904−6911.

Chapter 2, Page 47

2.2 Asymmetric Ring-Closing Metathesis of Protected Amines

2.2a Catalytic Asymmetric Ring-Closing Metathesis of Protected Amines. Our first goal was to study the desymmetrization reactions of differentially protected amines 2.12a-c (Scheme 2.2, see below).36 Our decision to study these protecting groups was twofold: (1) The acetate (2.12a), tert-butyl carbamate (2.12b), and urea (2.12c) groups would have orthogonal deprotection reaction conditions for the derived cyclic products. (2) The choice of protecting groups would provide insight into the functional group tolerance of chiral Mo complexes.

NMe

Me

Me

N

Me

Me

Me

2.12a R = Me2.12b R = NHPh

2.12c R = O-t-Bu

2.1470% conv., 75% ee, 63%

C6H6, 22 oC,24 h

O

Ar

OOMo

MePh

MeNAr

5 mol % 2.13Ar = 2,4,6,-i-Pr3(C6H2)

i-Pri-Pr N

Me

Me

Me

2.1563% conv., 29% ee

O NHPh

Scheme 2.2

N

Me

Me

Me

2.16>98% conv., 52% ee, 78%

O O-t-BuRO

MeO

To this end, we found Mo complex 2.13 to be the optimal catalyst in the conversion of trienes 2.12a-c to heterocycles 2.14, 2.15, and 2.16. We were gratified to find that a single catalyst was compatible with all three protecting groups; this allowed us to directly compare the outcome of all three catalytic reactions. Treatment of amines 2.12a-c with 5 mol % of Mo complex 2.13 delivers enantioenriched heterocycles 2.14 (70% conversion, 75% ee and 63% yield),37 2.15 (63% conversion and 29% ee), and 2.16 (>98% conversion, 52% ee and 78% yield) after 24 h. As the data in Scheme 2.2 illustrates, reactions to form acetate-protected heterocycle 2.14 and urea-protected heterocycle 2.15 proceed with comparable conversion to product (70% conversion and 63% conversion, respectively). The reaction to form tert-butyl carbamate-protected

(36) For the preparation of all metathesis substrates and metathesis substrate precursors please see the Experimental section 2.6. (37) The % conversion represents formation of product from consumption of substrate as monitored by 1H NMR of the unpurified reaction mixture. Undesired byproducts were typically not observed in these reactions. This holds true for all remaining conversions given in this thesis, unless otherwise mentioned.

Chapter 2, Page 48

heterocycle 2.16, however, proceeds to >98% conversion of desired product. This difference in conversion to product could be attributed to the tert-butyl carbamate serving as a better protecting group for nitrogen (in comparison to the acetate and urea groups) since it is sterically bulkier. Furthermore, the carbonyl oxygens of the amide and urea protecting groups are more nucleophilic, than the tert-butyl carbamate protecting group, likely favoring chelation to the metal center and decreasing catalyst activity.

5 mol % 2.2

OOMo

MePh

MeN

i-Pri-Pr

MeCF3

CF3

F3C

F3C Me

<2% conv.

N

O

MeMe

2.17

C6H6(2.5)

2.2b Evidence for Mo Complex Sequestration and Metathesis Deactivation.

Evidence for catalyst sequestration and deactivation comes from our observation of a chelated Mo alkylidene intermediate. During the course of our studies we found that in the presence of Mo complexes, the metathesis of amide-containing triene 2.17 did not proceed to appreciable levels of conversion to desired product (eq. 2.5). Even in the presence of achiral Mo complex 2.2, which is typically more reactive than chiral Mo complexes, < 2% of the desired product was formed. To garner insight into this lack of reactivity, we monitored this metathesis reaction by 1H NMR and observed formation of a new triplet (J = 4.7 Hz) at δ 11.97 ppm, corresponding to the chelated alkylidene intermediate shown in Figure 2.1. This notion of catalyst deactivation by chelation is also supported by the observation that ring-closing metathesis reactions of carbonyl containing substrates require longer reaction times than reactions with substrates lacking carbonyl moieties; reactions in Scheme 2.2 require 24 h, whereas reactions in eq. 2.2 and eq. 2.3 require 7 h and 3 h, respectively (see above).

H N

LnMo O

MeMe

Figure 2.1 Observation ofChelated Alkylidene

Chapter 2, Page 49

2.3 Asymmetric Ring-Closing Metathesis of Unprotected Secondary Amines

2.3a Studies of Sterically Encumbered Amine Metathesis Substrate. The next phase of our studies involved the enantioselective synthesis of amines that would contain alkyl substituents other than a methyl group at the nascent stereogenic center to be formed. When we prepared secondary amine 2.18, however, we encountered difficulties in protecting the nitrogen group (eq. 2.6, see below).38 Placing a phenyl group in place of a methyl group caused significant steric congestion adjacent to the nitrogen and impeded our attempts at protection of the amine. Based on this consideration, we decided to investigate the enantioselective synthesis of heterocycle 2.19; the idea was that the nitrogen would possibly be to sterically hindered to diminish the activity of Mo complexes.

NH

Ph

Me

Me

NH

Me

Me

PhMo Catalyst

(2.6)

2.18 2.19 2.3b Enantioselective Synthesis of an Unprotected Amine Heterocycle. Screening of

available chiral Mo complexes revealed Mo complex 2.5 to be the optimal catalyst for the desymmetrization of amine 2.18 (eq. 2.7). To this result, exposure of amine 2.18 to 5 mol % of Mo complex 2.5 delivers enantioenriched heterocycle 2.19 in 71% ee and 95% yield after 24 h. In addition, the reaction proceeds to >98% conversion of desired product. Encouraged by the efficient and selective synthesis of 2.19, we pursued the desymmetrization of related sterically hindered unprotected secondary amines.

NH

Ph

Me

Me

NH

Me

Me

Ph

2.18

Mo

N

MeMe

MeCl Cl

OO

C6H6, 22 oC,24 h

5 mol % 2.5

2.19>98% conv., 71% ee, 95%

(2.7)

(38) For the synthesis of 2.18, see the Experimental section 2.6.

Chapter 2, Page 50

As shown in Scheme 2.3 (see below), the next substrate under study was unprotected secondary amine 2.20, which contained a cyclohexyl group at the carbon adjacent to the amine (vs. phenyl in amine 2.18, eq. 2.7). Exposure of amine 2.20 to 5 mol % of Mo complex 2.21 affords heterocycle 2.22 in 33% ee and 56% yield after 24 h. Interestingly, this reaction only proceeds to 60% conversion of desired product. In studying the enantioselective synthesis of related heterocyle 2.23, where we replaced the cyclohexyl group with a hexyl group, the reaction proceeds to only 36% conversion to desired product. Unfortunately, we were unable to determine the enantioselectivity of 2.23.39 The asymmetric metathesis to form 2.22 is less selective than the reaction to form 2.19, pointing to the influence of substitution at the stereogenic center on catalyst enantioinduction.

NH

Cy

Me

Me

NH

Me

Me

Cy

2.20

Mo

N

PhMe

MeMe Me

OO

C6H6, 22 oC,24 h

5 mol % 2.21

2.2260% conv., 33% ee, 56%

Scheme 2.3

NH

Me

Me

2.2336% conv.

2.3c Influence of Substrate Substitution on Enantioselectivity. As discussed, in changing a phenyl to a cyclohexyl group, at the stereogenic center in heterocyclic products (2.19 in eq. 2.7 to 2.22 in Scheme 2.3), we observed decrease in the enantioselectivities of these products, begging the question “would a smaller substitution at the stereogenic center, such as a proton, lead to product formation in even lower enantioselectivity?” To address this question we studied the desymmetrization of secondary amine 2.24 (eq. 2.8, see below). Treatment of amine 2.24 with 5 mol % of Mo complex 2.25 delivers enantioenriched heterocycle 2.26 in 87% ee and 50% yield after 24 h; the reaction proceeds to 75% conversion of desired product. This result contradicted our initial hypothesis about the mechanistic details of this class of catalytic reactions.

(39) Attempts were made to separate enantiomers by chiral GC and HPLC. The enantiomers of amide and carbamate derivatives of 2.23 also did not separate by the above methods.

Chapter 2, Page 51

NH

H

Me

Me

NH

Me

Me

H

2.24

Mo

N

PhMe

Mei-Pr i-Pr

OO

C6H6, 22 oC,24 h

5 mol % 2.25

2.2675% conv., 87% ee, 50%

(2.8)

2.3d Proposed

Transition State Model for Asymmetric Ring-Closing Metathesis. From an initial analysis, it appears that no correlation can be made between the substitution pattern at the stereogenic center and enantioinduction. If a direct steric effect influences enantioinduction, we would expect 2.26 to form in lower enantioselectivity (vs. 2.19 in eq. 2.7 and 2.22 in Scheme 2.3, see above). After closer consideration, we devised a preliminary transition state model, shown in Figure 2.2, which serves to explain the enantioselectivities observed. It should be mentioned that further data is required to further support this model. Presumably, the Mo pre-catalyst initiates at the terminal olefin of the substrate molecule and then adopts a twist-chair conformation being either 2.27 or 2.28, depending on the substitution at the stereogenic center (shown in Figure 2.2, see below). This would be followed by coordination to one of the pendant enantiotopic olefins prior to metallacyclobutane formation. If the catalyst coordinates with the olefin illustrated by intermediate 2.27, the remaining alkyl-olefin containing side chain would occupy a psuedo-axial orientation and the phenyl group a pseudo-equatorial orientation. This would be highly preferred since a large size difference exists between the olefin side chain and the phenyl group. Relating this analysis to intermediate 2.28, where the

Mo

NR R

OO

Figure 2.2 Proposed Transition State Model

NH

Me

Ph

Me Mo

NR R

OO

NH

MeH

Me

2.27 2.28

Chapter 2, Page 52

stereogenic center would contain a proton instead of a phenyl group, it is likely that the olefin side chain occupies a pseudo-equatorial orientation. Intermediates 2.27 and 2.28 thus lead to products of appreciable enantioselectivity, 71% ee and 87% ee, respectively.

If the stereogenic center contains an alkyl group of similar size to the alkyl-olefin containing side chain, the difference in energy between the two possible enantiotopic transition states would be minimal and the product would likely form in low enantioselectivity; this is observed when the stereogenic center contains a cycohexyl group. As illustrated in Figure 2.3, the transition state energy of the diastereomeric transition states 2.29a and 2.29b would be close in value and the product would thus form in low enantioselectivity, as we do observe (33% ee). The difference in size between substituents at the stereogenic carbon and alkyl-olefin side chain can be reasoned by comparing “Winstein-Holness” A-values.40 If we take the A-value of the olefin-containing side chain to be close to that of an ethyl group (1.75 kcal/mol) and compare it to the A-value of a cyclohexyl group (2.15 kcal/mol) we see a difference of only ~0.40 kcal/mol. If we compare the difference in A-value energy between the olefin-containing side chain with that of a phenyl group (3.0 kcal/mol) we see a difference of 1.25 kcal/mol. Moreover, the difference in energy between the olefin-containing side chain and a proton is ~2.15 kcal/mol. Thus, as is observed, the product containing a proton at the stereogenic carbon (2.26 87%ee, eq. 2.8) is formed with greater enantioenrichment than the product containing a phenyl (2.19 71% ee, eq. 2.7), followed by the product containing the lowest A-value difference in the transition state, cyclohexyl (2.22 33% ee, Scheme 2.3).

Mo

NR R

OO

Figure 2.3 Proposed Transition State Model

NH

Me

Cy

Me Mo

NR R

OO

NH

MeCy

Me

2.29a 2.29b

(40) Hirsch, J. A. Top. in Stereochem. 1967, 1, 199−222.

Chapter 2, Page 53

2.4 Reactivity of Ru Complexes With Amine-Containing Substrates

2.4a Comparison of Ru Complexes for Asymmetric Ring-Closing Metathesis. Before concluding this chapter, it is important to emphasize the advantage of utilizing Mo complexes over Ru-based counterparts for this class of olefin metathesis transformations; the results in eq. 2.9 are illustrative. Treatment of amine 2.18 with 5 mol % of achiral Ru catalyst 2.30 in THF, and monitoring the reaction by thin layer chromatography after 5 h reveals that the desired product 2.19 does not form. Heating the reaction mixture to 65 oC, quenching the reaction after 12 h, does not promote formation of desired product. Analysis of the crude reaction mixture by 1H NMR analysis reveals that the reaction proceeds to only 27% conversion of desired product, whereas reaction with chiral Mo complex 2.5 proceeds to >98% conversion (compare with eq. 2.7, see above). Reaction of amine 2.18 with chiral Ru complex 2.31 is even less efficient, with isolation of starting amine 2.18 in >98% yield.

Ru

N

O

NX

Oi-Pr

Ph

Ru

N

Cl

N

Cl

Oi-Pr

Me

MeMe

Me

MeMe

Me

MeMe

NH

Ph

Me

Me

NH

Me

Me

Ph

2.18

THF, 65 oC,12 h

5 mol % 2.30

2.1927% conv.

(2.9)

2.31no reaction

2.5 Conclusions

These investigations discussed in this chapter resulted in the development of methods for the Mo-catalyzed enantioselective synthesis of heterocyclic amine compounds. These studies revealed that Mo complexes are sensitive to the substituents on nitrogen in olefin metathesis substrates with regards to efficiency and selectivity. Importantly, we were able to show for the first time that Mo complexes are tolerant of sterically hindered secondary amine substrates. These studies further demonstrated that Mo complexes are preferred for the olefin metathesis of amine containing substrates, over Ru-based counterparts.

Chapter 2, Page 54

2.6 Experimental Section41

NH (2.10)

2.32

O BMe

3

THF, 0 oC to 65 oC,12 h;

aq. NaOH, MeOH

NH

Ph

Me

Me

2.18 37% Synthesis of unprotected secondary amine 2.18. Amide 2.32 was prepared according to previously published literature procedures.42 The trimethallyl borane reagent was prepared (85% yield, 44 mmol) and purified according to a previously reported procedure used for the synthesis of triallyl borane; herein we employed BF3OEt2, methallyl chloride, and Mg turnings in Et2O.43 Amine 2.18 was prepared through a slight modification of a previously reported reductive bisalkylation of lactams as shown in eq. 2.10.44 A 15-mL round-bottom flask was charged with amide 2.32 (513 mg, 3.18 mmol), THF (1.00 mL), and the reaction vessel was cooled to 0 oC in an ice-bath. To this solution was added dropwise trimethyallyl borane (0.800 mL, 3.18 mmol), after which time the cooling was removed and the reaction mixture was allowed to warm to ambient temperature. At this time, the reaction vessel was fitted with a reflux condenser and the reaction mixture set to reflux at 65 oC. After allowing the reaction mixture to stir for 12 h, the heating bath was removed and the mixture was allowed to cool to ambient temperature. At this time, MeOH (3 mL) was added to the mixture and it was set to reflux a further 1 h, after which time the heating bath was removed and the resulting mixture was allowed to reach ambient temperature. At this point, a 5 M aqueous solution of NaOH (3 mL) was added to the mixture and allowed to stir for 30 min., after which time layers were separated, and the aqueous layer was washed with Et2O (3 x 20 mL). The combined organic layers were washed with H2O (20 mL), dried (Na2CO3), filtered, and the volatiles removed in vacuo. The resulting dark brown residue was purified by silica gel chromatography (CH2Cl2) to give 2.18 as yellow oil (0.300 g, 1.17 mmol, 37%). 1H NMR (400 MHz, CDCl3): δ 7.46−7.43 (m, 2H), 7.32−7.37 (m, 2H), 7.21−7.16 (m, 1H), 5.97 (tdd, J = 16.0, 10.0, 5.6 Hz, 1H), 5.25 (ddd, J = 17.2, 2.0, 1.6 Hz, 1H), 5.08 (41) For the spectral traces of compounds 2.12a, 2.24, 2.26, and 2.14 please see: Dissertation of Dr. Elizabeth Sattely. (42) “Synthesis of Cyclopropylpyrrolidines via Reaction of N-Allyl-N-propargylamides with a Molybdenum Carbene Complex. Effect of Substituents and Reaction Conditions,” Harvey, D. F.; Sigano, D. M. J. Org. Chem. 1996, 61, 2268−2272. (43) “Organoboranes. 44. A convenient, highly efficient synthesis of triorganylboranes via a modified organometallic route,” Brown, H. C.; Racherla, U. S. J. Org. Chem. 1986, 51, 427−432. (44) “A Convenient Synthesis of 2,2-Diallylated Nitrogen Heterocycles by Allylboration of Lactams,” Bubnov, Y. N.; Pastukhov, F. V.; Yampolsky, I. V.; Ignatenko, A. V. Eur. J. Org. Chem. 2000, 1503−1505.

Chapter 2, Page 55

(ddd, J = 10.4, 1.6, 1.2 Hz, 1H), 4.77 (td, J = 4.0, 1.6 Hz, 2H), 4.61 (s, 2H), 3.20 (td, J = 5.6, 1.2 Hz, 2H), 2.66 (d, J = 14.0 Hz, 2H), 2.52 (d, J = 14.0 Hz, 1H), 1.57 (s, 1H), 1.42 (s, 6H). 13C NMR (100 MHz, CDCl3): δ 145.9, 142.6, 137.4, 128.0, 127.0, 126.3, 115.2, 114.7, 61.1, 46.2, 45.1, 24.9.

NH (2.11)

2.33

Me

O BMe

3

THF, 0 oC to 65 oC,12 h;

aq. NaOH, MeOH

NH

Me

Me

Me

2.34 47% Synthesis of 2.34, the amine precursor to protected secondary amines 2.12a-c. Amine 2.34 was prepared in 47% yield by the same procedure used to synthesize amine 2.18 (eq. 2.10), however, allyl acetamide 2.33 was employed as the starting material for the reductive bisalkylation, shown in eq. 2.11. Protected amines 2.12a-c were prepared from secondary amine 2.34. 1H NMR (400 MHz, CDCl3): δ 5.91 (dddd, J = 11.6, 10.4, 5.6, 5.6 Hz, 1H), 5.17 (ddd, J = 17.2, 3.6, 1.6 Hz, 1H), 5.04 (ddd, J = 10.0, 2.8, 1.2 Hz, 1H), 4.97 (ddd, J = 4.0, 3.2, 1.6 Hz, 2H), 4.70−4.68 (m, 2H), 3.21 (ddd, J = 5.6, 1.2, 1.2 Hz, 2H), 2.16 (dd, J = 13.6, 13.6 Hz, 2H), 1.83 (s, 6H), 1.07 (s, 3H).

(2.12)NaH, THF,0 oC to 22 oC,

12 h

NMe

Me

Me

2.12a 44%

Cl Me

O

NH

Me

Me

Me

2.34

MeO

Synthesis of amide 2.12a. To a stirring solution of NaH (168 mg, 7.00 mmol) in 70 mL THF in a 100-mL round-bottom flask was added 2.34 (1.35 g, 7.00 mmol) by cannula transfer with a total of 10 mL THF. After allowing this mixture to stir for 1 h,the reaction vessel was cooled to 0 °C in an ice-bath. At this time, to this mixture was added acetylchloride (495 µL, 7.00 mmol) dropwise by syringe. The resulting mixture was allowed to stir for 12 h, at which point the reaction was quenched by the addition of H2O, (50 mL), and the resulting biphasic layers were separated. The aqueous layer was washed with Et2O (3 x 40 mL). The combined organic layers were dried (MgSO4), filtered, and the volatiles removed in vacuo. The resulting yellow residue was purified by silica gel chromatography (20:1 CH2Cl2:Et2O) to give 2.12a as colorless oil (725 mg, 3.08 mmol). IR (neat): 3087 (w), 2980 (m), 1671 (s), 1470 (m), 1401 (s), 1212 (m). 1H NMR (400 MHz, CDCl3): δ 5.79 (ddt, J = 17.2, 11.0, 5.5 Hz, 1H), 5.19−5.12 (m, 2H), 4.87 (dq, J = 2.5, 1.1 Hz, 2H), 4.68−4.67 (m, 2H), 3.83 (dt, J = 5.5, 1.8 Hz, 2H), 3.34 (d,

Chapter 2, Page 56

J = 13.2 Hz, 2H), 2.17 (d, J = 13.2 Hz, 2H), 2.07 (s, 3H), 1.73 (s, 6H), 1.32 (s, 3H). 13C NMR (100 MHz, CDCl3): δ 172.3, 143.0, 136.2, 116.3, 115.4, 63.3, 49.8, 45.8, 25.5, 25.0, 24.6. HRMS ES (m/z) Calcd for C15H25NO 236.2014 (M+H)+, Found 236.2011.

(2.13)NEt3, CH2Cl2,0 oC to 40 oC,

12 h

NMe

Me

Me

2.12b 80%

NH

Me

Me

Me

2.34

NHPhO

NO

Synthesis of amide 2.12b. A 25-mL round-bottom flask was charged with amine 2.34 (50.0 mg, 0.259 mmol), triethyl amine (36.0 µL, 0.259 mmol), CH2Cl2 (2.6 mL), and the reaction vessel was cooled to 0 oC in an ice-bath. To this solution was added phenyl isocyanate (28.0 µL, 0.259 mmol) by syringe, after which time the ice-bath was removed and the reaction mixture was allowed to reach ambient temperature. At this time, the reaction vessel was fitted with a reflux condenser and the reaction mixture set to reflux at 40 oC. After allowing the reaction mixture to stir for 12 h, the heating bath was removed and the mixture was allowed to cool to ambient temperature. At this time, CH2Cl2 (5 mL) followed by a saturated aqueous solution of NH4Cl (5 mL) was added to the reaction mixture. The resulting biphasic layers were separated, and the aqueous layer was washed with CH2Cl2 (3 x 20 mL). The combined organic layers were dried (MgSO4), filtered, and the volatiles removed in vacuo. The yellow residue was purified by silica gel chromatography (50:1 CH2Cl2:NEt3) to give 2.12b as yellow oil (65.0 mg, 0.208 mmol, 80%). 1H NMR (400 MHz, CDCl3): δ 7.60−7.25 (m, 4H), 7.02−6.98 (m, 1H), 5.94 (dddd, J = 17.2, 10.0, 5.2, 5.2 Hz, 1H), 5.45 (ddd, J = 17.2, 1.6, 1.6 Hz, 1H), 5.34 (ddd, J = 10.4, 1.6, 1.6 Hz, 1H), 4.92 (ddd, J = 4.4, 3.2, 1.6 Hz, 2H), 4.76−4.75 (m, 2H), 3.91 (ddd, J = 4.0, 2.0, 2.0 Hz, 2H), 3.40 (d, J = 12.8 Hz, 2H), 2.18 (d, J = 12.8 Hz, 2H), 1.81 (s, 6H), 1.38 (s, 3H).

THF,22 oC to 65 oC,

16 h 2.36 82%2.35

NH

Cy

Me

Me

2.20 65%

K2CO3, DMF,0 oC to 22 oC,

16 h

N ClMe

Mg(0), Br

H2NCy

Me

Me

Scheme 2.4

Chapter 2, Page 57

Synthesis of metathesis substrate 2.20. Metathesis substrate 2.20 was prepared through a slight modification of a known literature procedure, illustrated in Scheme 2.4.45 A 500-mL three-neck round-bottom flask, fitted with an addition funnel and reflux condenser, was charged with cyclohexanecarbonitrile 2.35 (3.30 mL, 27.5 mmol), magnesiu turnings (2.50 g, 103 mmol), and THF (200 mL). Methallyl chloride (6.80 mL, 68.7 mmol) and THF (68 mL) was slowly added to this solution by addition funnel, whilst heating the reaction vessel with a heat gun. After circa 10 mL of the methallyl chloride solution was added the mixture became olive green in color, signifying initiation of the reaction.46 At this moment, the reaction vessel was no longer heated with a heat gun and the remaining methallyl chloride solution was slowly added over a period of 1 h, at which point the reaction mixture was set to reflux at 65 oC. After 16 h, the heating bath was removed, the mixture was allowed to reach ambient temperature and then cooled to 0 oC in an ice-bath. At this point, the reaction was quenched by the addition of a saturated aqueous solution of NH4Cl (100 mL) and the resulting biphasic layers were separated. The organic layer was concentrated in vacuo to remove the majority of THF. The resulting mixture was diluted with Et2O (100 mL), washed with H2O (3 x 50 mL), dried (MgSO4), filtered, and the volatiles removed in vacuo. The resulting dark brown residue was purified by silica gel chromatography (20:1 CH2Cl2:MeOH) to give primary amine 2.36 as yellow oil (5.00 g, 22.6 mmol, 82%). 1H NMR (400 MHz, CDCl3): δ 4.92 (dd, J = 2.4, 1.2 Hz, 2H), 4.72 (dd, J = 2.4, 1.2 Hz, 2H), 2.20 (d, J = 13.2 Hz, 2H), 2.05 (d, J = 13.2 Hz, 2H), 1.81 (s, 6H), 1.79−1.6 (m, 2H), 1.40−1.00 (m, 9H). A 50-mL round-bottom flask was charged with amine 2.36 (0.50 g, 2.3 mmol), potassium carbonate (0.31 g, 2.3 mmol), DMF (13 mL), allyl bromide (0.20 mL, 2.3 mmol), and the resulting solution was allowed to stir for 16 h. At this time, the reaction mixture was diluted with Et2O (50 mL), washed with H2O (6 x 25 mL), dried (MgSO4), filtered and the volatiles removed in vacuo. The resulting brown residue was purified by silica gel chromatography (20:1 CH2Cl2:MeOH) to give secondary amine 2.20 as yellow oil (0.40 g, 1.5 mmol, 65%). 1H NMR (400 MHz, CDCl3): δ 5.90 (tdd, J = 18.0, 11.6, 6.0 Hz, 1H), 5.21 (ddd, J = 17.2, 3.6, 2.0 Hz, 1H), 5.16 (ddd, J = 17.2, 3.6, 2.0 Hz, 1H), 5.04 (dd, J = 3.6, 2.0 Hz, 1H), 5.01 (dd, J = 3.6, 2.0 Hz, 1H), 4.85 (dd, J = 2.4, 1.2 Hz, 2H), 4.75 (dd, J = 2.4, 1.2 Hz, 2H), 3.22 (m, 2H), 2.16 (s, 3H), 1.86 (s, 3H), 1.82−1.09 (m, 11H). 13C NMR (100 MHz, CDCl3): δ 144.4, 137.7, 114.6, 114.5, 60.6, 45.9, 42.9, 27.6, 27.5, 27.0, 25.5. (45) “Tertiary Alkyl Primary Amines, RR'R//CNH2. II,” Henze, H. R.; Allen, B. B.; Leslie, W. B. J. Am. Chem. Soc. 1943, 65, 87−89. (46) Caution should be taken as the reaction is highly exothermic at this point.

Chapter 2, Page 58

Representative procedure for tandem Mo-catalyzed Asymmetric Ring-Closing Metathesis; synthesis of amine 2.19. In a N2-filled glovebox, a 4-mL vial was charged with triene 2.18 (20.0 mg, 0.0783 mmol) and C6H6 (783 µL). To this solution was added 2.5 (3.2 mg, .0039 mmol) and the vial was fitted with a teflon lined cap. The reaction mixture was allowed to stir at 22 °C for 24 h, after which time the vial was removed from the glovebox and the volatiles were removed in vacuo. The resulting brown

residue was purified by silica gel chromatography (50:1 MeOH:CH2Cl2) (20.6 mg, 0.0744 mmol, 95%). IR (neat): 3383 (br), 2961 (m), 2917 (s), 2848 (m), 2369 (w). 1H NMR (400 MHz, CDCl3): δ 7.38−7.28 (m, 4H), 7.23−7.18 (m, 1H), 5.30 (s, 1H), 4.77 (s, 1H), 4.60 (s, 1H), 3.29 (d, J = 20.0 Hz, 1H), 3.03 (d, J = 20.0 Hz, 1H), 2.55−2.24 (m, 4H), 1.75 (s, 3H), 1.25 (s, 3H). 13C NMR (100 MHz, C6D6): δ 146.0, 142.8, 131.4, 128.4, 127.3, 126.8, 121.5, 115.2, 57.5, 51.9, 42.6, 39.2, 24.9, 24.0. HRMS ES (m/z) Calcd for C16H21N 228.1752 (M+H)+, Found 228.1747. Enantiomeric excess was determined by GLC analysis of the derived acetamide. Acetyl-protected 2.19-Ac: 1H NMR (400 MHz, CDCl3): δ 7.24−7.14 (m, 5H), 5.60 (s, 1H), 4.92 (s, 1H), 4.77 (s, 1H), 4.20−3.56 (m, 3H), 2.84 (t, J = 12.0 Hz, 2H), 2.18 (d, J = 16.0 Hz, 1H), 2.03 (s, 3H), 1.63 (s, 3H), 1.54 (s, 3H). The optical purity of 2.19-Ac was determined by GC analysis in comparison with authentic racemic material, shown below: β-dex chiral column, 130 oC, 15 psi, for a sample of 71% ee.

NH

Me

Me

Ph

2.19

Synthesis of amine 2.22. General procedure for ARCM was followed. Silica gel chromatography (1:1 Et2O:hexanes) was used for purification of 2.22, which was isolated as yellow oil (0.076 mmol, 56%). IR (neat): 2924 (m), 2855 (m), 1734 (w), 1639 (w). 1H NMR (400 MHz, CDCl3): δ 5.38 (s, 1H), 4.92 (s, 1H), 4.72 (s, 1H), 3.30 (dd, J = 41.6, 16.8 Hz, 2H), 2.32−1.98 (m, 4H), 1.84 (s, 3H), 1.66 (s, 3H), 1.48−0.87 (m, 11H). 13C

NMR (100 MHz, C6D6): δ 131.3, 127.9, 120.0, 114.6, 42.8, 42.0, 40.7, 36.1, 27.5, 27.4, 27.3, 27.2, 25.5, 24.1. HRMS ES (m/z) Calcd for C16H27N 234.2222 (M+H)+, Found 234.2224. Enantiomeric excess was determined by GLC analysis of the derived acetamide. Acetyl-protected 2.22-Ac: 1H NMR (400 MHz, CDCl3): δ 5.46 (s, 1H), 4.73 (s, 1H), 4.65 (s, 1H), 3.82−3.00 (m, 2H), 3.52−3.46 (m, 1H), 2.75−2.65 (m, 1H), 2.20−2.10 (m, 1H), 2.10 (s, 3H), 2.10−2.00 (m, 1H), 1.78 (s, 3H), 1.62 (s, 3H), 1.45−0.82

NH

Me

Me

Cy

2.22

Chapter 2, Page 59

(m, 11H). The optical purity of 2.22-Ac was determined by GC analysis in comparison with authentic racemic material, shown below: α-dex chiral column, 130 oC, 15 psi, for a sample of 33% ee.

Synthesis of amine 2.24. A reductive alkylation of allylformamide47 with trimethallylborane was performed according to the procedure for the synthesis of 2.34 (see eq. 2.11). Silica gel chromatography (4:1 CH2Cl2:hexanes washed with 1% v/v concentrated NH4OH) was used for purification of 2.24, which was isolated as yellow oil (8.5

mmol, 71%). The amine can be further purified by Kugelrohr distillation under vacuum to give colorless oil. IR (neat): 3087 (m), 2980 (m), 2936 (s), 1652 (m), 1457 (m), 1381 (m). 1H NMR (400 MHz, CDCl3): δ 5.85 (dddd, J = 16.8, 10.3, 6.2, 6.2 Hz, 1H), 5.15 (ddd, J = 17.2, 3.3, 1.5 Hz, 1H), 5.08 (d(br), J = 10.3 Hz, 1H), 4.81 (s(br), 2H), 4.75 (s(br), 2H), 3.26 (d, J = 5.9 Hz, 2H), 2.82 (dddd, J = 6.6, 6.6, 6.6, 6.6 Hz, 1H), 2.15 (dd, J = 13.9, 7.3 Hz, 2H), 2.07 (dd, J = 13.9, 6.2 Hz, 2H), 1.72 (s, 6H), 1.28 (s(br), 1H). 13C NMR (100 MHz, CDCl3): δ 143.5, 137.2, 116.0, 113.1, 52.0, 50.1, 43.4, 22.6. HRMS ES (m/z) Calcd for C12H21N 180.1752 (M+H)+, Found 180.1751.

NH

H

Me

Me

2.24

Synthesis of amine 2.26. General procedure for ARCM was followed. Silica gel chromatography (20:1 CH2Cl2:MeOH washed with 2% v/v concentrated NH4OH) was used for purification of 2.26, which was isolated as yellow oil (0.027 mmol, 50%, product is volatile). IR (neat): 3270 (w), 3075 (m), 2917 (s), 1659 (m), 1451 (s), 1381 (m), 890 (s). 1H NMR (400 MHz, CDCl3): δ 5.41 (s(br), 1H), 4.83 (s(br), 1H), 4.78 (s(br),

1H), 3.35 (s(br), 2H), 2.85 (dddd, J = 9.5, 7.7, 5.9, 4.4 Hz, 1H), 2.17−2.13 (m, 2H), 1.90−1.72 (m, 3H), 1.74 (s, 3H), 1.67 (s, 3H). 13C NMR (100 MHz, CDCl3): δ 142.9, 132.9, 120.1, 112.9, 50.4, 45.6, 45.3, 37.3, 23.4, 22.5. HRMS ES (m/z) Calcd for C10H17N 152.1439 (M+H)+, Found 152.1436. [α]D +7.7 (c = 1). Enantiomeric excess was determined by GLC analysis of the derived acetamide: 1H NMR (400 MHz, CDCl3,

NH

Me

Me

H

2.26

(47) Prepared according to literature procedure, see: “Acid-catalyzed cyclization reactions. IX. Formation of oxazolinium and thiazolinium cation from N-allyl and substituted N-allylamides, -urethans, -ureas, and –thioureas,” McManus, S. P.; Carroll, J. T.; Piiman Jr., C. U. J. Org. Chem. 1970, 35, 3768–3774.

Chapter 2, Page 60

25 C), ~2:3 mixture of amide rotamers: δ 5.39 (s(br), 0.4H), 5.33 (s(br), 0.6H), 5.10 (dd, J = 14.6, 7.3 Hz, 0.6H), 4.83 (s(br), 0.4H), 4.73 (s(br), 0.6H), 4.70 (s(br), 0.4H), 4.65 (s(br), 0.6H), 4.58 (d(br), J = 19.0 Hz, 0.4H), 4.08 (dd, J = 13.6, 6.6 Hz, 0.4H), 3.91 (d(br), J = 17.6 Hz, 0.6H), 3.73 (d(br), J = 17.6 Hz, 0.4H), 3.39 (d(br), J = 19.0 Hz, 0.4H), 2.38−2.20 (m, 2.4H), 2.11 (s, 1.2H), 2.10−2.07 (m, 0.6H), 2.04 (s, 1.8H), 1.93−1.79 (m, 1H), 1.78 (s, 1.8H), 1.77 (s, 1.2H), 1.71 (br s, 3H). The optical purity of 2.26-Ac was determined by GC analysis in comparison with authentic racemic material, shown below: CDGTA chiral column, 120 oC, 15 psi, for a sample of 87% ee.

Synthesis of amine 2.14. General procedure for ARCM was followed. Silica gel chromatography (15:1 CH2Cl2:Et2O) was used for purification of 2.14, which was isolated as colorless oil (0.140 mmol, 63%). IR (neat): 3068 (m), 2968 (s), 2923 (s), 2861 (m), 1646 (s), 1694 (s), 1243 (m), 1168 (m), 897 (s). 1H NMR (400 MHz, CDCl3): δ 5.53 (s(br), 1H), 4.83 (s, 1H), 4.66 (s, 1H), 3.85 (dd, J = 15.7, 4.8 Hz, 1H), 3.68 (dd, J = ), 3.02 (d, J = 13.6 Hz, 1H), 2.34 (d, J = 14.6 Hz, 1H), 2.30 (d, J = 13.6

Hz, 1H), 2.06 (s, 3H), 1.87 (d, J = 15.7 Hz, 1H), 1.74 (s, 3H), 1.49 (s, 3H). 13C NMR (100 MHz, CDCl3): δ 171.0, 143.3, 136.2, 117.6, 114.9, 58.7, 45.2, 44.8, 42.5, 26.9, 25.5, 24.2, 23.0. HRMS ES (m/z) Calcd for C13H21NO 208.1701 (M+H)+, Found 208.1698. The optical purity of 2.14 was determined by GC analysis in comparison with authentic racemic material, shown below: CDGTA chiral column, 140 oC, 15 psi, for a sample of

15.7, 1.5 Hz, 1H

75% ee.

N

Me

Me

Me

2.14

MeO

Synthesis of amine 2.15 (Scheme 2.2). General procedure for ARCM was followed. An optimized isolated yield was not obtained for amine 2.15. 1H NMR (400 MHz, CDCl3): δ 7.60−7.25 (m, 4H), 7.03−6.99 (m, 1H), 6.34 (br s, 1H), 5.47 (br s, 1H), 4.88 (ddd, J = 4.0,

Chapter 2, Page 61

2.8, 1.2 Hz, 1H), 4.74−4.73 (m, 1H), 4.0−3.80 (m, 2H), 2.76 (d, J = 14.0 Hz, 1H), 2.54 (d, J = 13.6 Hz, 1H), 2.29 (d, J = 15.6 Hz, 1H), 1.99 (d, J = 16.8 Hz, 1H), 1.80 (s, 3H), 1.76 (s, 3H), 1.25 (s, 3H). The optical purity of 2.15 was determined by HPLC analysis in comparison with authentic racemic material, shown below: Chiralpak AD (4.6 x 250 mm), 98:2 hexanes:i-PrOH, 1.0 mL/min, λ = 254 nm, for a sample of 29% ee.

N

Me

Me

Me

2.15

O NHPh

THF,22 oC to 65 oC,

16 h 2.38 15%2.37

NBoc

Me

Me

Me

2.12c 51%

KH, TBAI,THF, 22 oC,

12 h

Br

BocHNMe

Me

Me

Scheme 2.5

H2NMe

Me

Me Boc2O, K2CO3, DMAP

Synthesis of Metathesis Substrate 2.12c. Metathesis substrate 2.12 was prepared according to the procedure shown in Scheme 2.5. Amine 2.37 was prepared in 40% yield by the same procedure used to synthesize amine 2.36 (Scheme 2.4), however, acetonitrile was employed as the starting material for the bisalkylation. 1H NMR (400 MHz, CDCl3): δ 4.84 (br s, 2H), 4.64 (br s, 2H), 3.37 (br s, 4H), 2.04 (s, 3H), 1.73 (s, 6H), 1.46 (br s, 2H). A 250-mL round-bottom flask was charged with amine 2.37 (1.00 g, 6.52 mmol), potassium carbonate (1.35 g, 9.79 mmol), dimethylaminopyridine (39.8 mg, 0.326 mmol), and THF 65 mL). The reaction vessel was then fitted with a reflux condenser and the reaction mixture was set to reflux at 65 oC. After allowing the reaction mixture to stir for 16 h, the heating bath was removed, and the mixture was allowed to cool to ambient temperature. At this point, H2O (50 mL) was added to the mixture and the resulting biphasic layers were separated. The organic layer was concentrated in vacuo to remove a majority of the THF. The resulting mixture was diluted with Et2O, after which time it was washed with a saturated aqueous solution of NaCl (50 mL), H2O (50 mL), dried (MgSO4), filtered, and the volatiles removed in vacuo. The resulting brown residue was purified by silica gel chromatography (50:1 hexanes:Et2O) to provide tert-butyl carbamate 2.38 as colorless oil (256 mg, 1.01 mmol, 15%). 1H NMR (400 MHz, CDCl3): δ 4.90 (ddd, J = 3.6, 2.8, 1.2 Hz, 2H), 4.70−4.69 (m, 2H), 4.40 (br s, 1H), 2.77 (d, J =

Chapter 2, Page 62

13.6 Hz, 1H), 2.16 (d, J = 13.6 Hz, 1H), 1.80 (s, 6H), 1.43 (s, 9H), 1.17 (s, 3H). A 100-mL round-bottom flask was charged with 2.38 (256 mg, 1.01 mmol), potassium hydride (135 mg, 3.37 mmol), tetrabutylammonium iodide (18.7 mg, 0.0505 mmol), allyl bromide (0.300 mL, 3.45 mmol), and THF (20 mL). This mixture was allowed to stir for 12 h, after which time H2O (50 mL) was added to the mixture and the resulting biphasic layers were separated. The organic layer was concentrated in vacuo to remove a majority of the THF. The resulting mixture was diluted with Et2O, after which time it was washed with a saturated aqueous solution of NaCl (50 mL), H2O (50 mL), dried (MgSO4), filtered, and the volatiles removed in vacuo. The resulting brown residue was purified by silica gel chromatography (25:1 hexanes:Et2O) to provide metathesis substrate 2.12c as colorless oil (150 mg,0.0511 mmol, 51%). 1H NMR (400 MHz, CDCl3): δ 5.77 (dddd, J = 16.8, 10.0, 6.0, 6.0 Hz, 1H), 5.03 (dd, J = 17.2, 1.2 Hz, 1H), 4.97 (dd, J = 8.4, 1.6 Hz, 1H), 4.85 (br s, 2H), 4.65 (br s, 2H), 3.75 (d, J = 6.0 Hz, 1H), 3.20 (d, J = 12.8 Hz, 1H), 2.00 (d, J = 13.2 Hz, 1H), 1.72 (s, 6H), 1.44 (s, 9H), 1.29 (s, 3H). 13C NMR (100 MHz, CDCl3): δ 155.2, 142.9, 137.3, 115.3, 115.2, 114.9, 79.3, 61.6, 48.7, 46.9, 28.6, 24.7.

Synthesis of amine 2.16. General procedure for ARCM was followed. Silica gel chromatography (20:1 Et2O:hexanes) was used for purification of 2.16, which was isolated as colorless oil (0.0159 mmol, 78%). 1H NMR (400 MHz, CDCl3): δ 5.50 (br s, 1H), 4.84 (ddd, J = 4.0, 2.8, 1.6 Hz, 1H), 4.68−4.67 (m, 1H), 4.08−4.03 (m, 1H), 3.62−3.56 (m, 1H), 2.76 (d, J = 17.6 Hz, 1H), 2.28−2.21 (m, 3H), 1.91 (d, J = 16.0 Hz, 2H), 1.76

(s, 6H), 1.73 (s, 3H), 1.44 (s, 9H). 13C NMR (100 MHz, CDCl3): δ 155.8, 143.6, 134.5, 119.3, 115.3, 79.8, 57.4, 45.9, 44.0, 43.3, 29.0, 28.2, 24.6, 23.4. The optical purity of 2.16 was determined by GC analysis in comparison with authentic racemic material, shown below: CDGTA chiral column, 130 oC, 15 psi, for a sample of 52% ee.

N

Me

Me

Me

2.16

O O-t-Bu

Chapter 2, Page 63

Chapter 2, Page 64

Chapter 2, Page 65

Chapter 2, Page 66

Chapter 2, Page 67

Chapter 2, Page 68

Chapter 2, Page 69

Chapter 2, Page 70

Chapter 2, Page 71

Chapter 2, Page 72

Chapter 2, Page 73

Chapter 2, Page 74

Chapter 2, Page 75

Chapter 2, Page 76

Chapter 2, Page 77

Chapter 2, Page 78

Chapter 2, Page 79

Chapter 2, Page 80

Chapter 2, Page 81

Chapter 2, Page 82

Chapter 2, Page 83

Chapter 3, Page 84

Chapter Three

MO-CATALYZED ENANTIOSELECTIVE SYNTHESIS OF PIPERIDINES THROUGH ASYMMETRIC RING-OPENING/CROSS-METATHESIS

3.1 Introduction

Arguably, one of the most important reasons for researching organic chemistry is to develop methods that can be applied to the synthesis of agents of therapeutic value. Often times the inspiration for these methods are natural products that possess biological activity. Within this context, the next class of catalytic olefin metathesis reactions we developed was the asymmetric ring-opening/cross-metathesis (AROM/CM) of [3.2.1]-azabicycles for the enantioselective synthesis of 2,6-cis-disubstituted piperidines.

3.1a Piperidine Natural Products. Illustrated in Figure 3.1 are representative naturally occurring piperidine compounds that have demonstrated biological activity.

NMe

O OH

lobelineN

indolizidine 209D

Me

calvine

NH

MeC9H19

OH

dendrobate 241D

Figure 3.1 Bioglogically Active Piperidine Natural Products

N

OO

Fascinatingly, despite sharing similar structural architecture, these piperidine

natural products have been isolated from very dissimilar organisms. A common structural feature of these natural products is syn-substitution at the 2- and 6-position of the piperidine core. Calvine48 is a bicyclic piperidine that contains a seven-membered ring lactone and is isolated from coccinellid beetles belonging to the genus Calvia. This alkaloid is secreted when the beetles are disturbed and is a well-described process known

(48) “New Piperidine Alkaloids from Two Ladybird Beetles of the Genus Calvia (Coccinellidae),” Braekman, J-C,; Charlier, A.; Daloze, D.; Heilporn, S.; Pasteels, J.; Plasman, V.; Wang, S. Eur. J. Org. Chem. 1999, 1749−1755.

Chapter 3, Page 85

as “reflex bleeding”, which serves to protect the beetles from predators.49 Lobeline50 is an N-Me-containing piperidine isolated from lobelia inflata, which is generally found in eastern parts of North America and belongs to the lobelia family of alkaloids. Lobeline was originally referred to as Indian tobacco because the leaves of the plant were smoked by native Americans. Most recently, lobeline has been studied as a smoking deterrent and for the treatment of amphetamine addiction.51 Indolizidine 209D,52 belonging to the “bicyclic gephyrotoxin” family of alkaloids, is a 5-6-fused bicyclic piperidine that is isolated from the skin secretion of neotropical frogs. The indolizine alkaloids are a class of non-competative blockers for neuromuscular transmission.53 And although only one example is shown, dendrodate 241D, another common structural feature of piperidine alkaloids is an alcohol functionality at the 4-position of the piperidine; dendrotate 241D is isolated from the skin of poison dart frogs.54

3.1b Retrosynthetic Analysis of Dendrobate 241D; Substrates for AROM/CM. One set of substrates we studied for the enantioselective synthesis of piperidines were [3.2.1]-azabicycles, typified by 3.2 (Scheme 3.1). As shown in Scheme 3.1, the desymmetrization of an azabicycle such as 3.2 could be employed in the synthesis of piperidine alkaloids such as dendrobate 241D. Retrosynthetically, dendrobate 241D could be prepared from a 2,4,6-trisubstituted piperidine such as 3.1. Piperidine 3.1 in turn could be accessed from a Mo-catalyzed AROM/CM of azabicycle 3.2 with an appropriate olefin cross-partner 3.3. Although the free secondary alcohol is shown in azabicycle 3.2, initial studies involved different hydroxyl protected substrates since Mo complexes are not compatible with alcohol functionalities. Additionally, the functional group on the nitrogen of 3.2 is undefined

poison dart frog

(49) “Hemorrhage in a Coccinellid Beetle and Its Repellent Effect on Ants,” Happ, G. M.; Eisner, T. Science, 1961, 134, 329−331. (50) “History, chemistry and biology of alkaloids from Lobelia inflata,” Felpin, F-X.; Lebraton, J. Tetrahedron, 2004, 60, 10127−10153. (51) “Drugs to Fight Addictions,” Thayer, A. Chem. Eng. News 2006, 84, 21−44. (52) Daly, J. W.; Spande, T. F. In Alkaloids: Chemical and Biological Perspectives; Pelletier, S. W., Ed.; Wiley: New York, 1986; Vol. 4. Chapter 1. (53) Aronstam, R. S.; Daly, J. W.; Spande, T. F.; Narayanan, T. K.; Albequerque, E. X. Neurochemical Res. 1986, 11, 1227. (54) Daly, J. W.; Garraffo, H. M.; Spande, T. F. Amphibian Alkaloids. In The Alkaloids: Cordell, G. A., Ed.; Academic Press: New York, 1993; Vol. 43. pp. 185−289.

Chapter 3, Page 86

since it was apriori not possible to predict what N-containing functionalities would be most compatible with chiral Mo complexes.

NH

C9H19

OH

241D

NFG

OH

R

OH

FGN R+

AsymmetricRing-Opening/

Cross-Metathesis

Scheme 3.1

3.1 3.2 3.3 3.1c Retrosynthetic Analysis of Lobeline;

Azabicyclic Substrates for the Enantioselective Synthesis of 2,6-syn-Substituted Piperidines. In addition to studying the desymmetrization of azabicycles such as 3.2 (Scheme 3.1, see above), the desymmetrization of [3.2.1]-azabicycles such as trop-6-ene 3.5 (Scheme 3.2) was investigated. Retrosynthetically, the alkaloid lobeline could be accessed from an N-Me piperidine such as 3.4.

lobelia inflata

NMe

O OH

lobelineisolated from Lobelia inf lata

NMe

MeN +

AsymmetricRing-Opening/

Cross-Metathesis

Scheme 3.2

3.4

3.53.4

NMe

Interestingly, piperidine 3.4 could be accessed through a Mo-catalyzed

AROM/CM of trop-6-ene 3.5 with styrene. Two facets of this desymmetrization reaction merit discussion. First, the requisite cross-partner for this desymmetrization is styrene, an olefin donor that we knew was compatible with chiral Mo complexes. Thus, we had reason to be optimistic about the enantioselective reaction. Secondly, the desymmetrization substrate 3.5 would possess a tertiary amine, which based on our

Chapter 3, Page 87

previous experience (see Chapter 2) and literature reports was expected be compatible with Mo complexes.

3.2 AROM/CM approach to N-Me Piperidines

3.2a Initial Screen of Mo Catalysts for the Enantioselective Synthesis of N-Me Piperidines. Initial screening for the Mo-catalyzed AROM/CM of azabicycles with styrene is shown in Scheme 3.3.

2 equiv.styrene

C6H6, 22 oC, 12 h

5 mol %Catalyst

OTBS

MeN N

Me

OTBS

MeN

OTBS

NMe

OTBS

3.7 70%>98% E:Z

3.8 30%

Scheme 3.3

+

3.6

020406080

100

% c

onv.

/% e

e ee

A B C D E F G H

3.7

3.8

% ee 3.7

Catalysts Screened

R R

Mo

N

R'Me

Me

OO

A R = i-Pr, R' = PhB R = Cl, R' = MeC R = Me, R' = Ph

O

Ar

OOMo

MeR'

MeN

RR

Ar

Mo

N

PhMe

Me

OO

H

D R = i-Pr, R' = PhAr = 2,4,6,-i-Pr3(C6H2)

O

Ar

OOMo

MePh

MeN

i-Pri-Pr

Ar

F Ar = 2,4,6,-i-Pr3Ph(C6H2)G Ar = CHPh2

O

Ar

OOMo

MePh

MeN

CF3

Ar

E Ar = 2,4,6,-i-Pr3(C6H2)

Chapter 3, Page 88

The substrate we first decided to study was azabicycle 3.6. By studying this particular azabicycle we were able to address the issues of: (1) Preparing piperidines with an alcohol functionality at the 4-position and (2) the feasibility of preparing enantioenriched N-Me-containing piperidines. To this end, a screen of available chiral Mo catalysts revealed that two complexes, C and H, are capable of providing the desired piperidine 3.7. Chiral Mo complex C, however, delivers piperidine 3.7 in racemic form, were as chiral Mo complex H affords piperidine 3.7 in 94% ee an in >98% E:Z olefingeometry.55 The dimeric byproduct 3.8 is also formed during the course of the reaction. The isolated yields in Scheme 3.3, 70% yield for 3.7 and 30% yield for 3.8, are the results obtained with Mo complex H. At this point, our next objective was to optimize the reaction conditions for the enantioselective synthesis of 3.7 and garner insight into the mechanism for this class of desymmetrization reactions.

3.2b Optimization and Mechanistic Studies into the Mo-catalyzed AROM/CM of Azabicycles. Based on previous studies concerning the Mo-catalyzed desymmetrization of norbornenes, we presumed that a Mo benzylidene i (Scheme 3.4, see below) performs the initial ring-opening of azabicycle 3.6 to provide the Mo alkylidene intermediate ii. From here, reaction of ii with styrene would deliver the desired product if the Molybdenum alkylidene reacts with another equivalent of styrene. Alternatively, reaction of ii with another azabicyclic molecule of 3.6 would provide a Mo alkylidene iii, which would lead to alkylidene iv, and ultimately result in the formation of the dimer observed. This mechanistic proposal would be the most likely scenario if azabicyle 3.6 is strained enough to compete with styrene toward reaction with Mo alkylidene 3.9. To better gauge the reactivity of this azabicyle, we performed a reaction of 3.6 with Mo complex H in the absence of styrene and monitored the outcome by 1H NMR analysis. To this end, exposure of azabicycle 3.6 to 5 mol % of Mo complex H in

Figure 3.2 polymerizationof azabicycle 3.6 within 30 min

OOMo

MePh

MeN

i-Pri-Pr

MeCF3

CF3

F3C

F3C Me

deuterated benzene results in a rapid ring-opening metathesis polymerization (ROMP) reaction. As judged by 1H NMR analysis, 75% of the initial azabicycle is polymerized within 30 min. As a comparison, when the reaction is performed with the

(55) All piperidine products in this chapter were isolated with >98% E:Z olefin geometry as determined by 1H NMR analysis; large J-coupling values, indicative of trans olefins, were found for all styrenyl olefinic protons in piperidine products. For a discussion of this concept, see: Silverstein, R. M.; Webster, F. X. In Spectrometric Identification of Organic Compound; Rose, N., Swain, E., Eds.; John Wiley & Sons, Inc. New York, 1998;

Chapter 3, Page 89

achiral Schrock catalyst (Fig. 3.2), azabicycle 3.6 is fully-consumed and polymerized within 30 min.

Scheme 3.4OTBS

N

OTBS

MeN

LnMo

MeN

6 6

substrate polymerization

N

NMe

OTBS

i i

MeLnMo

OTBS

NMe

OTBS

no styreneC D , 22 oC, 30 min

OTBS

MeN

iv 75% conversion to

i i

3.6

ii i

3.6

NMe

OTBS

MeN

LnMo

OTBS n

Mo

N

PhMe

Me

O

5 mol % H

Mo PhO

iOTBS

MeN

3.6

LnMoO

O

One approach to solving the problem of dimer formation in the desymmetrization of azabicycle 3.6 was to increase the equivalents of styrene in the reaction (eq. 3.1, see below). During our initial screen of chiral catalysts for the AROM/CM of 3.6 we employed 2 equivalents of styrene. From previous experience, we knew that stillbene formation in the presence of Mo complexes is slow.56 This means that Mo methylidene complexes, that are typically detrimental to ring-openin cross-metathesis reactions, (56) “Catalytic Asymmetric Ring-Opening Metathesis/Cross Metathesis (AROM/CM) Reactions. Mechanism and Application to Enantioselective Synthesis of Functionalized Cyclopentanes,” La, D. S.; Sattely, E. S.; Ford, J. G.; Schrock R. R.; Hoveyda A. H. J. Am. Chem. Soc. 2001, 123, 7767−7778.

Chapter 3, Page 90

should not be formed in appreciable levels. Gratifyingly, after screening a variety of styrene equivalents in the desymmetrization of 3.6, we found that utilizing 10 equivalents of styrene provided the best results for this transformation. Treatment of azabicycle 3.6 with 5 mol % Mo complex H in the presence of 10 equivalents of styrene affords piperidine 3.7 in 94% ee and 95% yield after 1h. These optimized reaction conditions were then employed to study the scope of Mo-catalyzed AROM/CM reactions of a variety of azabicycles with various electronically and sterically modified styrene cross-partners.

NMe

MeN

10 equiv.styreneC6H6,

22 oC, 1 h 3.7 94% ee, 95%

OTBS OTBS

(3.1)

3.6

Mo

N

PhMe

Me

OO

5 mol % H

3.2d Influence of Electron Poor and Electron Rich Styrenes on the Reaction Rate

and Enantioselectivity of AROM/CM Reactions. Among one of the most interesting features of this class of catalytic enantioselective olefin metathesis reactions is the effect that the aryl-olefin cross-partner has on the outcome of the transformation. Specifically, we found that in AROM/CM reactions of azabicycles with electron rich and electron poor

product within 1 h and 54% conversion within 4 h. Notably, the product is isolated in

styrene cross-partners the piperidine products were formed in different rates and selectivities. First, we studied the AROM/CM of azabicycle 3.6 with 2 equivalents of p-CF3-styrene in the presence of 5 mol % H (eq. 3.2, see below) and monitored the conversion of the reaction by 1H NMR analysis. Two equivalents of p-CF3-styrene were initially used because we wanted to conduct a direct comparison with the results we observed during our initial screen in the metathesis of 3.6 with styrene (Scheme 3.3, see above). To this end, treatment of azabicycle 3.6 with 5 mol % Mo complex H in the presence of 2 equivalents of p-CF3-styrene proceeds to 26% conversion to the desired product 3.9 within 30 min. The reaction proceeds to 35% conversion to the desired

only 64% ee, as opposed to 94% ee when the olefin cross-partner is styrene.

Chapter 3, Page 91

NMe

MeN

C6H6, 22 oC

3.9 64% ee30 min 26% conv.1 h 35% conv.4 h 54% conv.

OTBS OTBS

(3.2)

CF3

CF3

2 equiv.

MeOO

5 mol % H+

3.6

Mo

N

Ph

Me

The results obtained in the AROM/CM of azabicycle 3.6 with p-CF3-styrene are in stark contrast to the results observed when an electron rich olefin, p-OMe-styrene, is used as an olefin cross-partner. As shown in eq. 3.3, under identical reaction conditions to those used in eq. 3.2, desired piperidine 3.10 is formed at a faster rate and in greater enantiopurity. Treatment of azabicycle 3.6 with 5 mol % of Mo complex H in the presence of 2 equivalents of p-OMe-styrene proceeds to 50% conversion to the desired product 3.10 within 30 min. The reaction proceeds to 50% conversion to the desired product within 1 h and >98% conversion within 4 h. A proposal for this difference in reactivity follows.

NMe

MeN

6 6o

OTBS OTBS

(3.3)OMe

Mo

N

PhMe

Me

O

5 mol % H+

3.10 89% ee20 min 50% conv.1 h 80% conv.

C H , 22 C

OMe

2 equiv.3.6

4 h >98% conv. Prior to presenting the following mechanistic proposal it should be noted that there is no definitive evidence for this hypothesis, rather it is an attempt to explain the results observed by way of empirical evidence. As illustrated in Figure 3.2 (see below),

O

Chapter 3, Page 92

Mo benzylidene intermediates such as i are highly reactive for olefin metathesis reactions because the benzylidene carbon possesses significant δ− character and in addition the highly Lewis acidic nature at the metal center (+6) facilitates the coordination of olefins. The drawback to such reactivity is exemplified by what would happen to Mo benzylidene i in the presence of an alcohol; the benzylidene carbon would undergo protonatation and the complex would possess an additional alkoxy ligand resulting in formation of ii. Now, if we apply these principles to benzylidenes 3.11 and 3.12, previously discussed in equations 3.2 and 3.3, we can perform an analysis to explain their differences in reactivity with azabicycle 3.6 for AROM/CM reactions. A benzylidene such as 3.11a, formed from reaction of Mo complex H with p-CF3-styrene (see eq. 3.2 above), would decrease the electron density at the benzylidene carbon and therefore decrease electron density at the metal center making it more Lewis acidic and consequently more likely to

the benzylidene carbon would possess more electron density, and thus be more reactive. Benzylidene 3.12a can be viewed as a destabilized alkylidene by the repulsive effect of having two δ− carbon atoms adjacent to each other. Furthermore, if we consider Mo alkylidene iii, the initial product from the ring-opening metathesis of 3.6 with Mo complex 3.12a, we can see that de-stabilization at the benzylidene carbon would be reduced. In short, it appears that, at least in Mo-catalyzed AROM/CM of azabicycles, electron poor benzylidenes lead to decrease metathesis activity whereas electron rich benzylidenes lead to increased metathesis activity. This mechanistic proposal relies on the rate-determining step of the AROM/CM reaction to be initial ring-opening of the strained azabicyle, rather than formation of the subsequent metallacyclobutane and its break down to generate the piperidine product. However, it is still not clear to us what the rate-determining step is in these reactions.

bind an olefin. This can be more easily seen in 3.11b, a resonance form of 3.11a. Since the benzylidene carbon would now donate electron density into the aromatic ring it would be expected to be less reactive in olefin metathesis reactions, and this is what we observed in the AROM/CM shown in eq. 3.2 above. Another way of rationalizing the inherent reactivity of this type of Mo benzylidene is by viewing 3.11a as a stabilized, and therefore less reactive benzylidene. This electronic effect would be reversed in the case of a Mo benzylidene such as 3.12a, formed from reaction of Mo complex H with p-OMe-styrene (see eq. 3.3 above); this effect being more evident in 3.12b, a resonance form of 3.12a. In this case,

Chapter 3, Page 93

Mo

N

O

CF

MoO

3.11a

3.12a

δ+

δ−

+6Mo

N

PhO

i

δ− Mo

N

O

CF3

3.11b

δ−

δ

ROH Mo

N

O

ii

OR

Ph

Mo

N

O

OMe

3.12b

δ−

δ−

OTBS

N

O

O

3

N

O

OMe

Figure 3.2 Electronics of Different Alkylidenes

NMe

OMe

Mo

iii

O O

O

+

O

O

Chapter 3, Page 94

As we have seen the electronics of aryl olefins plays a crucial role in the rates of olefin metathesis reactions. When we employ optimized reaction conditions, however, we observe that these electronic effects are moot since reactions are complete within the same reaction time (eq. 3.4). To this end, exposure of azabicycle 3.6 to 5 mol % Mo complex H, in the presence of p-CF3-styrene (10 equiv.) or p-OMe-styrene (10 equiv.) delivers the desired piperidines 3.9 in 64% ee (88% yield) and 3.10 in 89% ee (92% yield) within 1h. Since increasing cross-partner equivalents clearly results shorter reaction times these results may suggest that the rate-determing step in these reactions is release of the Mo alkylidene formed after the initial ring-opening metathesis of azabicycles.

It should be mentioned that at this time it is difficult to put forth a model accounting for the difference in enantioselectivities observed between these two reactions. One explanation, however, is that benzylidenes 3.11 and 3.12 (see Fig 3.2 above) exist in different syn/anti ratios, and thus different selectivity levels; further mechanistic studies are required to support this hypothesis.

NMe

OTBS

NMe

OTBS

OMe

3.10 89% ee, 92%

3.9 64% ee, 88%

CF3

MeN

10 equiv.cross-partner

C6H6,22 oC, 1 h

OTBS

(3.4)

Mo

N

PhMe

Me

OO

5 mol % H

3.6

3.2e Scope of Aromatic Cross-Partners with N-Me Azabicycles. The next phase

of our studies involved expanding the scope of our newly developed method; employing the optimized reaction conditions with Mo complex H,57 we were able to form various enantioenriched N-Me piperidines (Figure 3.3, see below). Piperidine 3.14 is isolated in 98% ee and 91% yield. Of interest in the formation of this piperidine is that o-Me-styrene is used as the olefin cross-partner in the metathesis. At the outset we were not

(57) Reaction conditions involved 1 equiv. of the requisite azabicycle, 5 mol % Mo complex H and 10 equiv. olefin cross-partner in a 1.0 M solution in C6H6 at 22 oC. All reactions were stopped aft r 1 h. e

Chapter 3, Page 95

su the metathesis activity of Mo complexes would be inhibited by o-substituted styrenes. Piperidine 3.14 is isolated in 88% ee and 86% yield. Notably, this piperidine is contains an aryl bromide that can be potentially employed in various metal-catalyzed cross-coupling reactions.58 To show that orthogonal hydroxyl protecting groups can be employed in this class of transform

re if

16 is isolated in 92% ee and 90% yield. As was mentioned in the introdu

iously mentioned alkaloids, for example lobeline.

ations we utilized an azabicyclic substrate with a benzyl protecting group as opposed to a silyl ether. To this end, piperidine 3.15 is isolated in 95% ee and 86% yield. To demonstrate the stereochemistry at the hydroxy functionality is not detrimental to the asymmetric reaction we employed an azabicyclic substrate with opposite stereochemistry at the oxygenated carbon. To this result, piperidine 3.

ction of this chapter, we directed our studies toward the synthesis of 2,4,6-substituted piperidines in addition to the enantioselective synthesis of piperidines lacking substitution at the 4-position. To this end, piperidine 3.17 is isolated in 80% ee and 64% yield.59 Unfortunately, in this reaction piperidine 3.17 is isolated in moderate selectivity because this piperidine shares many structural features of prev

3.14 88% ee, 86%3.13 98% ee, 91%

NMe

OTBS

NMe

OTBS

BrMe

NMe

OBn

NMe

OTBS

NMe

3.15 95% ee, 86% 3.16 92% ee, 90% 3.17 80% ee, 64%

Figure 3.3 Scope of N-Me Piperidine Products

(58) (a) “Palladium-Catalyzed Cross-Coupling Reactions of Organoboron Compounds,” Miyaura, N.; Suzuki, A. Chem. Rev. 1995, 95, 2457−2483. (b) “The Heck Reaction as a Sharpening Stone of Palladium Catalysis,” Beletskaya, I. P.; Cheprakov, A. V. Chem. Rev. 2000, 100, 3009−3066. (c) “Selected Patented Cross-Coupling Reaction Technologies,” Corbet, J.-P.; Mignani, G. Chem. Rev. 2006, 106, 2651−2710. (d) “Aryl-Aryl Bond Formation by Transition-Metal-Catalyzed Direct Arylation,” Alberico, D.; Scott, M. E.; Lautens, M. Chem. Rev. 2007, 107, 174−238. (59) The reason for the low isolated yield of 3.17, in comparison to other piperidines in Fig. 3.3, is that we found 3.17 to be somewhat volatile.

Chapter 3, Page 96

3.3 Mo-catalyzed AROCM of Carbamate-Protected Azabicycles

3.3a Enantioselective Synthesis of Cbz-protected Piperidines. Next, we turned our attention on the desymmetrization of carbamate-protected azabicycles. One key issue we wanted to address, besides studying different substitution at nitrogen, was whether it would be feasible to perform the ring-opening of sp2 hybridized azabicycles. The synthesis of carbamate-protected azabicycles, discussed in the experimental section of this chapter, have one interesting caveat; these substrates are prepared by ring-closing metathesis reactions of 2,6-divinyl piperidines.60 These details gave us reason to believe ring-opening reactions of carbamate-protected azabicycles might not be feasible.

To our initial skepticism, we found Mo complexes to be compatible with Cbz-protected azabicycles for the enantioselective synthesis of piperidines (eq. 3.5).

Mo

N Me

NCbz

OTBSOTBS

CbzN 10 equiv.

styreneC6H6,

22 oC, 24 h 3.19 90% ee, 93%>98% conversion

PhMeO

O

10 mol % H

3.18

(3.5)

Treatment of azabicycle 3.18 with 10 mol % Mo complex H in the presence of 10

equivalents of styrene affords piperidine 3.19 in 90% ee, 93% yield, and >98% conversion after 24 h. At this point, a comparison is merited between this reaction and the AROM/CM reactions of N-Me azabicycles. First, we found the optimal catalyst loading in formation of 3.19 to be 10 mol % as opposed to the catalyst loadings of 5 mol % used in the desymmetrization of N-Me azabicycles; reactions to form 3.19 with 5 mol % of Mo complex H proceeded to 45% conversion of 3.19 after 24 h.61 Second, despite performing the reaction to form 3.19 with 10 mol % of Mo Complex H, the reaction is

(60) A detailed discussion of the reasons why the ring-closures of carbamate-protected piperidines is included in the following sub-section. (61) Additionally, The formation of 3.19 with 5 mol % and 10 mol % H was monitored by 1H NMR analysis. With 5 mol % of H, piperidine 3.19 is formed in <5% conversion after 1 h, ~20% conversion

after 3 h, and ~56% conversion after 6 h. after 3 h, and ~30% conversion after 6 h. With 10 mol % of H, piperidine 3.19 is formed in ~25% conversion after 1 h, ~40% conversion

Chapter 3, Page 97

complete within 24 h, as opposed to a reaction time of 1 h for the synthesis of N-Me

ed was azabicy

azabicycles. Based on these differences, it is likely that carbamate-protected azabicycles are much less strained than their N-Me counterparts. An additional point of interest in these enantioselective metathesis reactions is that Mo complex H delivers the same antipode of piperidine products; N-Me piperidines and N-Cbz piperidines are formed with the same absolute stereochemistry. This was verified by chemical correlation; the carbamate of 3.19 was reduced to deliver N-Me piperidine 3.7 (see eq. 3.1 above) and was found to have the same sense of enantioinduction.

3.3b Reversibility of Cbz-protected Piperidine. The next substrate we studicle 3.20, a diastereomer of azabicycle 3.18, and results are summarized in Scheme

3.5.

NCbz

OTBS

3.21 85% ee, 80%85% conversion

OTBS

CbzN

OTBS OBn

Ph

OBn

10 equiv.styreneC6H6,

22 oC, 24 h

Mo

N

PhMe

Me

OO

10 mol % H

3.20

3.21b

CbzN

10 mol % H

10 equiv.

C H ,

N+

1,3

Scheme 3.5

N

O

OBnN O

TBSO H

H

Ph

N OTBSO H

HMo Cat.

3.21a2,6-equitorial 2,6-axial

3.20

OTBS OTBSOTBS

A strain

styreneCbzN

Cbz6 6

22 oC, 24 h 3.20 ~15% formed 3.21 85% ee3.21 85% ee

Chapter 3, Page 98

Treatment of azabicycle 3.20 with 10 mol % of Mo complex H in the presence of 10 equivalents of styrene furnishes Cbz-protected piperidine 3.21 in 85% ee, 85% conversion, and 80% yield after 24 h. A feature of this reaction that quickly struck our attention was that the reaction did not proceed to completion, in contrast to the metathesis of 3.18 (see eq. 3.5 above). In light of this observation we raised the following question: is it possible that this metathesis reaction is reversible by way of the Mo catalyst performing a ring-closing metathesis of piperidine 3.21? It is well documented that the preferred conformation of sp2 hybridized piperidines with substituents at the 2- and 6-

e 3.5 (see above), the sp2 character of the nitrogen group causes severe A1,3-strain which is relieved when the piperidine adopts a conformation such as 3.21b. From this analysis we can see that the olefins at the 2- and 6-position of this piperidine would be in perfect alignment to undergo a metallacyclobutane formation in the presence of a Mo catalyst and re-form the starting azabicycle 3.20. To explore this possibility we subjected the isolated piperidine 3.21 to the AROM/CM reaction conditions. Indeed, we found that Mo complex H is capable of performing a ring-closing metathesis of 3.21 in roughly 15% conversion with no change in its initial enantiopurity, and in addition delivers azabicycle 3.20 (as judged by 1H NMR and chiral HPLC analysis).

3.3c Scope of Enantioselective Synthesis of Cbz-protected Piperidines. Our subsequent objectives involved studying the scope of enantioenriched Cbz-protected piperidine products we could form through Mo-catalysis, illustrated in Figure 3.4 (see below). In metathesis reactions to form these piperidines we employed the optimized reaction conditions previously discussed.63 First, we performed the AROM/CM reaction of a Cbz-protected azabicyclic substrate lacking a hydroxyl functionality.

position is that in which the substituents are positioned in a pseudo-axial and not pseudo-equatorial orientation.62 These reports have demonstrated that in piperidines, such as 3.21a in Schem

(62) (a) “Allylic strain in six-membered rings,” Johnson, F. Chem. Rev. 1968, 68, 375−413. (b) “The “A” value for the methyl group in 2-methylpiperidines,” Fraser, R. R.; Grindley, T. B. Tetrahedron Lett. 1974, 15, 4169−4172. (c) “Synthesis of 2,6-diacetonylpiperidine. X-ray diffraction analysis of its N-benzoyl derivative,” Quick, J.; Mondello, C.; Humora, M.; Brennan, T. J. Org. Chem. 1978, 43, 2705−2708. (63) Reaction conditions involved 1 equiv. of the requisite azabicycle, 10 mol % Mo complex H and 10 equiv. olefin cross-partner in a 1.0 M solution in C6H6 at 22 oC. All reactions were stopped after 24 h.

Chapter 3, Page 99

NCbz

NCO2Et

Figure 3.4 Scope of Carbamate Piperidine Products

NCbz

OTBS

NCbz

OTBS

OMe

3.25 80% ee, 65% 3.26 86% ee, 92%

Me

3.24 80% ee, 60%

NCO2Et

OTBS

3.22 80% ee, 82% 3.23 90% ee, 72%

To this end, piperidine 3.22 is isolated in 80% ee and 82% isolated yield. We

then moved on to study the desymmetrization of an ethyl-carbamate protected azabicycle to see how a different smaller carbamate-protecting group would effect the selectivity of the reaction. This change in the protecting group on nitrogen proved beneficial and we found the metathesis reaction to be more selective with the smaller carbamate. Ethyl-carbamate protected piperidine 3.23 is isolated in 90% ee and 72% yield (vs. 80% ee in 3.22). Next, we performed the AROM/CM on an azabicycle containing a silyl ether. Similar

nically and sterically modified aryl aromatic cross-partners. Piperidine 3.25, generated with the use of p-OMe-styrene as olefin-cross partner in the asymmetric reaction, is isolated in 80% ee and 65% isolated yield. Finally, with the use of o-Me-styrene as the olefin cross-partner, the catalytic enantioselective reaction delivers piperidine 3.26 in 86% ee and 92% yield.

3.4 Mo-catalyzed AROM/CM of Bn-protected Azabicycles

3.4a Low Selectivity in the Enantioselective Synthesis of N-Bn Piperidines. In our

to our first carbamate-protected azabicyclic substrate (see 3.18 in eq. 3.5 above), but with the smaller ethyl-carbamate protecting group. To this result, piperidine 3.24 is isolated in 80% ee and 60% yield. The results of this reaction proved that the smaller protecting group does not always lead to products of greater selectivity; compare 3.19 of 90% ee in eq. 3.5 with 3.24 of 80% ee in Fig. 3.4 (see above). Additional points of mention for this class of AROM/CM reactions include the use of electro

continuing efforts to understand the limits and breadth of enantioenriched piperidine products accessible through olefin metathesis, we investigated the desymmetrization of Bn-protected azabicycles (eq. 3.6, see below).

Chapter 3, Page 100

NBn

BnN

10 equiv.styreneC6H6,

22 oC, 1 h 3.28 21% ee, 85%

OTBS OTBS

Mo

N

PhMe

Me

OO

10 mol % H(3.6)

3.27 Unlike previous desymmetrization reactions that delivered N-Me and Cbz-

protected piperidine products in high enantioselectivity, reactions that provided Bn-protect

3.5 Mo-catalyzed AROM/CM of Azabicycles with Non-Aryl Olefin Cross-Parnters

pursued reactions involving non-aryl olefin cross-partners. During these studies we found that symm

on the

ed amines were only slightly selective. Treatment of azabicycle 3.27 with 10 mol % of Mo complex H in the presence of 10 equiv. of styrene furnishes piperidine 3.28 in only 21% ee and 85% yield after 1 h. The origin of this major dissimilarity in selectivity remains unclear. One possibility is that the Lewis basic electrons on nitrogen (in the case of N-Me azabicycles) or electrons of the oxygens (in the case of the carbamate-protected azabicycles) serve to direct the Mo complex during the enantiodetermining step of the metathesis reaction. In the case of N-Bn-protected azabicycles, the size of the Bn group might diminish the ability of the lone pair electrons on nitrogen to effectively direct the Mo complex during the enantiodetermining step of the metathesis reaction. Despite these observations being a limitation of Mo-catalyzed enantioselective olefin metathesis reactions, they provided impetus for us to study Ru catalysts for this class of transformations (discussed in Ch 5).

3.5a Mo-catalyzed AROM/CM Reactions of Alkyl Olefin Cross-partners. After studying AROM/CM reactions of aryl-olefin cross-partners we N

a etric reactions with non-aryl olefin cross-partners were generally not as selective and efficient as reactions with aryl-olefins. Illustrated in Scheme 3.6 (see below) are studies we performed

3.35

MoOO

Figure 3.5 Mo Methylidene

Chapter 3, Page 101

AROM/CM of azabicycle 3.29 with alkyl olefins. To this end, exposure of 3.29 to 5 mol % Mo complex H and 10 equiv. vinylcyclohexane delivers the desired piperidine 3.30 in 20% ee, 43% yield, and 95% conversion (consumption of starting material as judged by 1H NMR analysis) after 14 h. And treatment of azabicycle 3.29 with 5 mol % of Mo complex H in the presence of 10 equiv. of hexene delivers the desired piperidine 3.33 in <5% ee, 25% yield, and 80% conversion (consumption of starting material as judged by 1H NMR analysis) after 14 h. A reason for the low yields in these reactions is the formation of meso byproducts arising from a bis-cross metathesis of the olefin cross partner, 3.31 in the case of vinyl cyclohexane and 3.34 in the case of hexene. The other meso byproduct from the reactions, the 2,6-divinyl piperidine 3.32, presumably forms from a ring-opening of azabicycle 3.29 with Mo methylidene intermediate 3.35, shown in Figure 3.5. Mo methylidene 3.35 forms through the homodimerization of either vinylcylohexene and hexene; these homodimers are observe by 1H NMR analysis of the reaction mixture. These undesired side reactions are likely responsible for the low enantioselectivity observed in piperidines 3.30 and 3.33. It should be noted that the alkyl olefin cross partners would be more readily consumed through dimmerization if less than 10 equiv. of alkyl cross-partners are employed in these reactions. In addition, Mo methylidene intermediates have been shown to undergo decomposition pathways, explaining why these reactions do not proceed to complete consumption of 3.29.

Me

N

PhMe

Me

O

5 mol % H

Scheme 3.6

OTBS OTBSOTBS

NMe

+ NMe

RC6H6,

+

%

NR

3.31 R = Cy3.32 R = H

%

Me

NMe

RR

3.34 R = hex3.32 R = H

o

Mo

3.29

+

OTBS

O

3.30 20% ee, 4395% conv.

10 equiv.

NMeC6H6,

22 oC, 14 h

5 mol % H

3.33 <5% ee, 2580% conv.

+N

10 equiv.

22 C, 14 h

3.29

OTBSOTBS

Chapter 3, Page 102

3.5b Mo-catalyzed AROM/CM Reactions of Miscellaneous Olefin Cross-partners. Moving our attention away from aryl and alkyl olefin cross-partners, we explored the potential of conducting AROM/CM reactions of azabicycles with enol-ether olefin cross-partners (eq. 3.7).

NCO2Et

3.36 70%~1:1

EtO

OEt

+N

OOEt

OOMo

MePh

MeN

i-Pri-Pr

MeCF3

3

3

F3C Me

C6H6,22 oC, 12 h

2 equiv.3.29

(3.7)

After screening of available Mo complexes we found that only the achiral

Schrock Mo complex was compatible with ethyl vinyl ether for the ring-opening cross-metathesis of azabicycle 3.29. Thus, treatment of azabicycle 3.29 with 5 mol % of the Mo Schrock catalyst in the presence of 2 equiv. of ethyl vinyl ether delivers the desired piperidine product in 70% yield, as a 1:1 mixture of E- and Z-olefin isomers, withi

CF

F C

5 mol %

n 12 h. This result underlines the need for the development of more active chiral Mo complexes and provides potential future areas of study in asymmetric ring-opening cross-metathesis reactions. Other olefin cross-partner investigated that proved incompatible with either chiral or non-chiral Mo complexes include vinyl trimethoxysilane, vinyl pinocal borane, allyl pinacol borane, and acrylonitrile.

3.6 Conclusions

This chapter disclosed the first examples of Mo-catalyzed asymmetric ring-opening metathesis/cross-metathesis reactions for the enantioselective syntheses of piperidines. It was shown that N-Me- and carbamate-containing azabicycles were compatible with Mo complexes for the desymmetrization reactions. These investigations also demonstrated that a variety of aryl-olefin cross-partners can be employed for catalytic enantioselective reactions. Additionally, these studies provided insight into the mechanism of this class of ring-opening/cross-metathesis reactions. Specifically, it was

ation of Mo benzylidene complexes is required prior to the initial ring-pening reactions of azabicycles. It was also shown that azabicycles exhibit different

reactivities towards electron poor and electron rich aryl-olefins. Furthermore, it was

shown that formo

Chapter 3, Page 103

shown

to the potential application of this method to natural product synthesis. The shortcomings of this method, namely in the synthesis of N-Bn piperidines and cross-metathesis reactions with alkyl olefins underline the need to continue developing novel Mo complexes for asymmetric olefin metathesis.

3.7 Experimental Section64

Synthesis and spectral data for azabicyclic substrates that deliver 2,4,6-substituted piperidine products. For the preparation of azabicyclic substrates that provide 2,4,6-substituted piperidines, we utilized intermediate azabicycle 3.37, which has been previously prepared.65

that certain desymmetrization reactions are in equilibrium and that Mo complexes can perform the ring-closing of piperidine products. Importantly, the products formed bear close resemblance to a variety of biologicially naturally occurring alkaloids, pointing

O

L-Selectride®

THF,

OH

CbzN

3.37

−78 oC to 22 oC,24 h

CbzN

3.38 81%single diastereomer

(3.8)

®

o

Preparation of azabicycle 3.38 (eq. 3.8). A two-neck 200 mL round-bottom flask, fitted with addition funnel, was charged with 3.37 (1.00 g, 3.89 mmol), THF (40 mL), and cooled to −78 °C in a dry ice/acetone bath. L-Selectride (a 1.00 M solution in THF, 4.30 mL, 4.28 mmol, 1.1 equiv.) was added to this solution over 10 min by addition funnel, after which time the mixture was warmed to 22 C, and the resulting mixture was allowed to stir for 1 h. At this period, the reaction mixture was cooled to 0 °C in an ice bath, and the reaction was quenched by the (slow and sequential) addition of a 1.0 M

(20 mL), and a solution of H2O2 (37 wt % in H2O, 20 mL). The solution of NaOHresulting mixture was warmed to 22 oC, and was allowed to stir for 15 min. At this time, EtOAc (20 mL), and a 1.0 M solution of HCl (20 mL) were added to this mixture. The resulting mixture was transferred to a 250-mL separatory funnel, and the resulting biphasic layers were separated. The aqueous layer was washed with CH2Cl2 (3 × 100 mL). The combined organic layers were dried (MgSO4), filtered, and the volatiles were

(64) In some instances, rotamers of carbamate- and amide-containing azabicycles were observed as determined by variable temperature 1H NMR analysis. Additionally, this is likely the reason that in some instances 13C NMR spectra do not appear to have the correct number of carbon peaks. In these instances the 13C NMR spectra is reported as observed. (65) “Synthesis of Bridged Azabicyclic Structures via Ring-Closing Olefin Metathesis,” Neipp, C. E.; Martin S. F. J. Org. Chem. 2003, 68, 8867−8878.

Chapter 3, Page 104

removed in vacuo. The resulting yellow oil was purified by silica gel chromatography (3:2 EtOAc:hexanes) to provide 3.38 as colorless oil and as a single diastereomer (0.811 g, 3.12 mmol, 81%). 1H NMR (400 MHz, CDCl3):66 1:1 mixture of carbamate rotamers, δ 7.37−7.31 (m, 5H), 6.43−6.30 (m, 2H), 5.17 (s, 2H), 4.67−4.60 (m, 2H), 3.97−3.91 (m, 1H), 2.34−2.14 (m, 2H), 1.82−1.76 (d, J = 14.8 Hz, 2H).

Preparation of azabiycle 3.18. Azabicycle 3.18 was prepared by TBS protection of the alcohol in azabicy

OTBS

cle 3.38.67 The relative stereochemistry of oxyl group carbon was determined by correlation to nOe

MR (400 MHz, CDCl3), 1:1 mixture of carbamate rotamers:68 δ 7.37−7.26 (m, 5H), 6.04−5.99 (m, 2H), 5.18 (s, 2H), 4.68−4.65 (m, 2H), 3.99 (t, J = 5.2 Hz, 1H), 2.25−2.00 (m, 2H), 1.60−1.52 (m, 2H), 0.81 (s, 9H), −0.02 (s, 6H). 13C NMR (100 MHz, CDCl3): δ 152.7, 137.1, 134.0, 133.6, 128.6, 128.0, 127.9, 66.7, 65.1, 57.4, 36.0, 35.2, 25.8, 17.9, −4.8. HRMS ES (m/z) Calcd for C21H31NO3Si 373.2073 M+

the protected hydrN

analysis of azabicycle 3.20, the diastereomer of 3.18 (see below). To further verify the identity of these diastereoisomers, a nOe experiment of 3.18 was

also performed. IR (neat): 2951 (m), 2920 (m), 2858 (m), 1710 (s), 1412 (m) 1300 (m), 1244 (m), 1089 (s). 1H N

3.18

Cbz

Found 373.2071.

The azabicyclic precursor to 3.24 (shown here on the left) was prepared by utilizing the same sequence to form 3.18, however, ethylchloroformate was used in the place of CbzCl at the beginning of the synthetic route (This azabicycle was isolated as colorless oil and in comparable yields to those in the preparation of azabicycle 3.18). IR

(neat): 2955 (m), 2923 (m), 2848 (w), 1702 (s). 1H NMR (400 MHz, CDCl3), 1:1 mixture of carbamate rotamers:68 δ 6.18−6.02 (m, 2H), 4.59−4.44 (m, 2H), 4.16 (q, J = 7.2 Hz, 2H), 3.99 (t, J = 5.6 Hz, 1H), 2.22−2.02 (m, 2H), 1.56 (d, J =14.4 Hz, 2H), 1.26 (t, J = 7.2 Hz, 3H), 0.86 (s, 9H), −0.03 (s, 6H). 13C NMR (100 MHz, CDCl3): δ 134.1, 133.6, 65.1, 60.9, 57.3, 57.2, 35.9, 35.1, 25.8, 17.9, 15.0, −4.8. HRMS ES (m/z) Calcd for C16H29NO3Si 311.1917 (M)+, Found 311.1923.

N OEt

OTBSO

precursor topiperidine 3.24

(66) Residual water is present in the 1H NMR spectra of this compound.

8) Residual TBSOH is present in the 1H NMR spectra of this compound.

(67) “Protection of Hydroxyl Groups as tert-butyldimethylsilyl Derivatives,” Corey, E. J.; Venkateswarlu, A. J. Am. Chem. Soc. 1972, 94, 6190−6191. (6

Chapter 3, Page 105

90%

CbzN

3.38

THF, 65 oC,1 h

OTBS

MeN

3.6

(3.9)LiAlH4

OTBS

Synthesis of azabicycle 3.6 (eq. 3.9, see above). Representative procedure for lithium aluminum hydride reduction of carbamate protected azabicycles: In a N -filled glovebox, a 4-mL vial was charged with 3.18 (40.0 mg, 0.0800 mmol), THF (0.800 mL), and lithium aluminum hydride (6.1 mg, 0.16 mmol, 2 equiv.). The vial was tightly sealed with a Teflon cap, placed in a heating mantle at 65 C, and the reaction mixture was allowed to stir for 1 h. At this point, the reaction mixture was cooled to 22 C over 15 min, the vial was re oved from the glovebox, and the reaction was stopped by the slow and sequential addition of H O (40.0 µL, 1 µL/mg of 3.18), a 3.8 M solution of NaOH (0.120 mL, 3 µL/g of 3.18) and H O (40.0 µL, 1 µL/mg of 3.18. The resulting mixture was subjected to vacuum filtration (to remove Al salts), and eluted with EtOAc. The filtrate was dried (Na SO ), re-filtered, and the volatiles were removed in vacuo. The resulting yellow residue was purified by silica gel chromatography (9:1 CH2Cl2:MeOH) to afford 7 as colorless oil (18.2 mg, 0.0720 mmol, 90%). IR (neat): 2930 (s), 2854 (m), 1256 (w), 1067 3

3.34 (br s, 2H),−0.04 (s, 6H). 1

3

−5.1. HRMS E 14 27

2

o

o

m2

2

2 4

(s). 1H NMR (400 MHz, CDCl ): δ 5.90 (s, 2H), 3.95 (t, J = 5.6 Hz, 1H), 2.27 (s, 3H), 2.13−2.07 (m, 2H), 1.57 (d, J = 14.0 Hz, 2H), 0.84 (s, 9H), 3C NMR (100 MHz, CDCl ): δ 131.4, 65.9, 64.2, 41.4, 37.1, 25.5, 17.6, S (m/z) Calcd for C H NOSi 253.1857 (M)+, Found 253.1862.

CbzN

3.37

O

i-PrOH, THF,65 oC,12 h

OH

CbzN

3.39 60%

(3.10)II

Synthesis of Azabicycle 3.39, precursor to azabicycle 3.20 (3.10). In a N2-filled glovebox, a 25-mL round-bottom flask was charged

Sm (0),

with Sm (132 mg, 0.880 mmol), and 1,2-diiodoethane (248 mg, 0.880 mmol), after which the flask was capped with a rubber septa. The flask was then removed from the glovebox, at which point THF (4.5 mL) was added to the mixture. After the resulting solution became deep blue (circa 1 h), a solution of 3.37 (0.100 g, 0.440 mmol) in isopropanol (33.0 µL, 0.440 mmol), and THF (4.0 mL) was added to the deep blue solution. At this moment, a reflux condenser was

o

oled to 22 oC, and the reaction was quenched by the attached to the flask, and the reaction mixture was allowed to stir at 65 C for 12 h. At this time, the reaction mixture was co

Chapter 3, Page 106

addition of H2O (10 mL). The resulting mixture was filtered through celite by vacuum filtration, and eluted with EtOAc. The volatiles were removed from the filtrate in vacuo, the resulting mixture was transferred to a 60-mL separatory funnel, and the mixture was washed with CH2Cl2 (3 x 20 mL). The combined organic layers were dried (MgSO4), filtered, and concentrated in vacuo to yield yellow oil. This oil was purified by silica gel

3

3

17 3

Azabicycle 3.20 was prepared by TBS protection of the alcohol in azabicycle 3.39. The relative stereochemistry at the protected hydroxyl group carbon was determined by nOe analysis. IR (neat): 2951 (m), 2920 (m), 2858 (m1412 (m) 1300 (m), 1244 (m), 1089 (s). H NMR (400 MHz,

CDCl ), 1:1 mixture of carbamate rotamers: 7.37 7.26 (m, 5H), 6.04 5.99 (m, 2H), 5.18 5.10 (m, 2H), 4.68 4.65 (m, 2H), 3.94 3.90 (m, 1H), 1.85 1.77 (m, 2H), 1.72 1.50 (m, 2H), 0.86 (s, 9H), s . C NMR (100 MHz, CDCl ): 152.2, 136.9, 131.1, 130.8, 128.5, 128.3, 128.0, 127.9, 66.7, 65.5, 65.4, 57.3, 57.1, 35.0, 34.9, 34.2, 25.8, 22.3,

4.6. HRMS ES (m/z) mass Calcd for C21H31NO3Si 373.2073 (M)+ Found

The azabicyclic precursor to 3.16 (shown on the left) was prepared by lithium aluminum hydride reduction of the carbam te in 3.20; for a

chromatography (dry load, 1:1 to 1:2 hexanes:EtOAc to 100% EtOAc) to give 3.39 as colorless oil (68.0 mg, 0.260 mmol, 60%). IR (neat): 3404 (br), 2945 (w), 2926 (w), 1685 (s), 1418 (s) 1319 (m), 1288 (m), 1102 (s). 1H NMR (400 MHz, CDCl ), 1:1 mixture of carbamate rotamers: δ 7.37−7.26 (m, 5H), 6.04−5.99 (m, 2H), 5.18−5.10 (m, 2H), 4.68−4.65 (m, 2H), 3.90 (dddd, J = 12.8, 12.8, 6.6, 6.4 Hz, 1H), 1.97 (dd, J = 12.8, 6.4 Hz, 2H), 1.86 (s, 1H), 1.64−1.58 (m, 1H), 1.51−1.45 (m, 1H). 13C NMR (100 MHz, CDCl ): δ 152.3, 136.7, 131.0, 130.8, 128.5, 128.1, 128.0, 66.8, 64.7, 57.2, 34.7, 33.9. HRMS ES (m/z) Calcd for C15H NO 259.1208 (M)+, Found 259.1206.

), 1710 (s), 1

3 δ − −− − − − −

0.00 ( , 6Η) 133 δ

CbzN

3.20 H

OTBS

nOe

OTBS

N OBn

O

18.1, −373.2068.

MeN

OTBS

arepresentative procedure, see azabicycle 3.6 (eq. 3.9, see above). IR (neat): 2930 (s), 2855 (m), 1470 (w), 1256 (m), 1099 (s). 1H NMR (400 MHz, CDCl3): δ 5.87 (s, 2H), 3.74 (dddd, J = 13.6, 8.8, 6.8, 6.8 Hz, 1H),

3.47 (br s, 2H), 2.22 (s, 3H), 1.90−1.85 (m, 2Η) 1.70−1.60 (m, 2Η), 0.82 (s, 9H), 0.00 (s, 6H). 13C NMR (100 MHz, CDCl3): δ 129.5, 66.5, 65.4, 40.5, 35.8, 25.9, 18.1, −4.4. HRMS ES (m/z) Calcd for C14H27NOSi 253.1862 (M)+, Found 253.1866.

precursor topiperidine 3.16

Chapter 3, Page 107

The azabicyclic precursor to 3.15 (shown on the left) was prepared by benzyl protection69 of the alcohol in azabicycle 3.38 (eq. 3.8, see above), followed by lithium aluminum hydride reduction of the carbamate; for a representative procedure, see azabicycle 3.6 (eq 3.9, see above). IR 2930 (s), 2848 (m), 1451 (w), 1067 (s). 1H NMR (400 MHz, CDCl3, TMS): δ

7.34–7.25 (m, 5H), 6.04 (s, 2H), 4.40 (s, 2H), 3.64 (dd, J = 6.4, 6.4 Hz, 1H), 3.41 (br s, 2H), 2.31 (s, 3H), 2.17–2.11 (m, 2H), 1.86 (d, J = 14.0 Hz, 2H). 13C NMR (100 MHz, CDCl3): δ 139.5, 132.1, 127.3, 127.2, 71.6, 70.1, 66.0, 53.6, 41.8, 33.7. HRMS ES (m/z) Calcd for C15H19NO 229.1467 (M)+, Found 229.1461. Synthesis and spectral data for azabicyclic substrates that deliver 2,6-substituted piperidine products. Preparation of the azabicyclic precursor to piperidine 3.23 (shown below) was previously reported from azabicycle 3.42 (vide infra).70 However, the cost of th

MeN

OBn

precursor topiperidine 3.15

e commercially available starting materiaof the reported synthetic route prompted us to develop an alternative route to azabicycle 3.42, shown in Scheme 3

l (6-exo-hydroxytropinone, $200/g from Aldrich) and the length

.7 (see below).

O O

OHHO +

i) 3 N HCl, 4 h;6 N NaOH

ii) NaOAc,allylamine22 oC, 18 h

2 4130 oC, 1 h

ii) KOH, 130−180 oC3 h

3.40unpurified

3.41unpurified

O

HN

OH

N

OH

OO OMeMeO i) N H , EtOH,

Scheme 3.7

78% average per step

o

i) 1 equiv. NEt ,O

CH2Cl2 3.4234% overall0 oC 22 C

4 h

31 equiv. ClCO2Et

ii) 1 equiv. NEt31 equiv. MsCl

3.41

HN

OH

N

OMs

OEt

(69) “Nouvelle Methode de Benzylation D'hydroxyles Glucidiques Encombres,” Czernecki, S.; Georgoulis, C.; Provelenghiou, C. Tetrahedron Lett. 1976, 17, 3535−3536 (70) “Synthesis of (±)-Epibatidine and Its Analogues,” Bai, S.; Xu, R.; Chu, G.; Zhu, X. J. Org. Chem. 1996, 61, 4600−4606.

Chapter 3, Page 108

A modified prdimethoxyfuransubsequent steintermediate 3.4pot procedure w

be utilized to access different

arbinol.

40 mL) was added, and the resulting solution was allowed to stir for 30 min. At this period, the mixture was added to a solution of NaOAc (98.4 g, 1.20 mol, 8 equiv.), allylamine (22.5 mL, 0.300

mol, 2 equiv.), and acetonedicarboxylic acid (43.8 g, 0.300 mol, 2 equiv.) in de-ionized H2O (2.0 L) in a 4-L Erlenmeyer flask. The reaction mixture was allowed to stir for 18 h. To this mixture was added K2CO3 (25 g, 1.25 wt %), NaCl (25 g, 1.25 wt %), and the resulting mixture was allowed to stir for 1 h, after which time it was transferred to a 4-L separatory funnel, and washed with CH2Cl2 (6 × 500 mL). The combined organic layers were dried (MgSO4), filtered, and concentrated in vacuo to afford 3.40 as viscous brown oil (20.0 g); 3.40 was used in the subsequent reaction without purification. IR (neat): 3408 (br), 2948 (m), 1709 (s), 1413 (m), 1344 (m). 1H NMR (400 MHz, CDCl3, TMS): δ 5.95 (dddd, J = 16.4, 12.8, 10.4, 6.4 Hz, 1H), 5.32−5.25 (m, 2H), 4.07 (dd, J = 6.8, 2.4

Hz, 1H), 3.69 (br s, 1H), 3.52−3.40 (m, 5H), 2.68−2.56 (m, 2H), 2.26−1.94 (m, 2H), 1.80 (br s, 1H). 13C NMR (100 MHz, CDCl3): δ 208.4, 135.3, 117.3, 74.9, 66.1, 56.9, 51.2, 44.4, 42.1, 40.7. HRMS ES (m/z) Calcd for C10H15NO2 182.1181 (M)+, Found 182.1175.

Synthesis of azabicycle 3.41. A 500-mL round-bottom flask was charged

ocedure for the Robinson-type annulation71 of allylamine with 2,5-, and acetonedicarboxylic acid afforded 3.40, which was used in the p without purification. Wolff-Kischner reduction of 3.40 provided 1, which was also taken forward without purification. A two-step, one-as then performed to obtain azabicycle 3.42; a single purification was

required in the sequence carried out to arrive at 3.41. To this end, treatment of 3.41 with ethylchloroformate followed by treatment with mesyl chloride provided pure 3.42 after silica gel chromatography (34% overall yield).72 It is worth noting that an advantage of our synthesis for 3.42 is that intermediate 3.41 canazabicyclic derivatives; the amino group in 3.41 can be selectively functionalized prior to derivatization of the c

Synthesis of azabicycle 3.40. Dimethoxyfuran (20.0 g, 0.150 mol) was dissolved in a 3.0 N aqueous solution of HCl (280 mL) in a 500-mL Erlenmeyer flask, and was allowed to stir for 4 h. At this time, a 6.0 N aqueous solution of NaOH (1

N

OH

O

3.40

HN

OH

3.41

(71) “Chemical Synthesis and Pharmacology of 6- and 7-Hydroxylated 2-Carbomethoxy-3-(p-

00, 43, 3283−3294.

ion to an enocyclic olefin.

tolyl)tropanes: Antagonism of Cocaine's Locomotor Stimulant Effects,” Zhao, L.; Johnson, K. M.; Zhang, M.; Flippen-Anderson, J.; Kozikowski, A. P. J. Med Chem. 20(72) The stereochemistry at the carbon containing a mesyl group of 3.42 was not determined as the next step involves an eliminat

Chapter 3, Page 109

with azabicycle 3.40 (20.0 g, 0.110 mol), anhydrous EtOH (200 mL), hydrazine hydrate (48.0 mL, 0.990 mol, 9 equiv.), the flask was fitted with a reflux condenser, and the mixture was allowed to stir at 120 oC for 1.5 h. At this time, EtOH was removed in vacuo to yield viscous dark brown oil. To this oil was added powdered KOH (56.0 g, 0.990 mol, 9 equiv.),73 and the resulting mixture was allowed to stir at 130 oC for 1 h, at 160 oC for 1 h, and at 180 oC for 2.5 h. After this period, the reaction mixture was cooled to 22 oC, and the reaction was quenched by the addition of H2O (200 mL). The resulting mixture was transferred to a 1-L separatory funnel, and washed with CH2Cl2 (6 × 500 mL). The combined organic layers were dried (Na2SO4), filtered, and concentrated in vacuo to give 3.41 as viscous brown oil (14.0 g); 3.41 was used in the next step without purification. IR (neat): 3361 (

O

N

OMs

OEt

3.42

H NMR (400 MHz, CDCl3): 5.48 (br s, 2H), 4.25−4.20 (m, 1H), 3.85−3.80 5 (br s, 1H), 2.20−2.15 (m, 1H), 1.90−1.82 (m, 1H), 1.80−1.65 (m, 2H), , 2H), 1.40−1.20 (m, 2H). 13C NMR (100 MHz, CDCl3): δ 73.7, 63.5, 55.5, 7.6, 16.8. HRMS ES (m/z) Calcd for C7H14NO 128.1075 (M+H)+, Found

Synthesis of azabicycle 3.42. A two-neck 250-mL round-bottom flask, fitted with an addition funnel, was charged with 3.42 (2.20 g, 15.9 mmol), CH2Cl2 (100 mL), Et3N (2.22 mL, 15.9 mmol, 1 equiv.), and cooled to 0 oC in an ice bath. Ethyl chloroformate (1.51 mL, 15.9 mmol, 1 equiv.) was added to this solution dropwise by addition funnel; after the addition was complete the ice bath was removed, and the mixture was allowed to stir for

2 h. To this solution was added Et3N (2.22 mL, 15.9 mmol, 1 equiv.), and the resulting solution was cooled to 0 oC in an ice bath. To this solution was added methanesulfonyl chloride (1.23 mL, 15.9 mmol, 1 equiv.) dropwise by addition funnel; after the addition was complete the ice bath was removed, and the mixture was allowed to stir for 2 h. At this time, the reaction was quenched by the addition of H2O (75 mL), the mixture was

to a 250-mL separatory funnel, and the resulting biphasic layers were The organic layer was washed with a saturated aqueous solution of NaCl (1 x d H2O (1 x 75 mL), after which time the organic layer was dried (MgSO4), d the volatiles were removed in vacuo. The resulting yellow oil was purified el chromatography (gradient elut

br), 3261 (br), 2926 (s), 2864 (m), 1635 (w), 1536 (w), 1437 (m). 1

(m, 1H), 3.21.62−1.42 (m39.7, 29.5, 2128.1071.

transferredseparated. 75 mL), anfiltered, anby silica g ion, 1:1 Et2O:hexanes followed by 100% Et2O) to prdimethoxyfuran). Compound 3.42 has been prepared previously; the physical and spectral data for 3.42 were identical to those previously reported.65 1H NMR (400 MHz,

ovide 3.42 as colorless oil (2.78 g, 9.00 mmol, 34% yield starting from 2,5-

(73) A mortar and pestle were used to crush KOH granules into a powder.

Chapter 3, Page 110

is azabicycle has been prepared previously and full characterization data for were reportedpreviously reδ 6.02−5.99 (1.25 (t, J = 7.

CDCl3): δ 5.13 (dd, J = 6.4, 3.2 Hz, 1H), 4.50−4.35 (m, 2H), 4.16 (q, J = 7.2 Hz, 2H), 3.03 (s, 3H), 2.27−2.17 (m, 1H), 1.78−1.23 (m, 7H), 1.20 (t, J = 7.2 Hz, 3H).

Syntheis of the azabicyclic precursor to 3.23 (shown on the left). A 100-mL round-bottom flask was charged with 3.42 (7.21 g, 23.3 mmol), collidine (30 mL), and 1,8-diazabicyclo[5.4.0]undec-7-ene (4.20 mL, 27.9 mmol, 1.2 equiv.). The flask was fitted with a reflux condenser, and the reaction mixture was allowed to stir at 160 oC for 24 h. At this

period, the reaction mixture was cooled to 22 oC, and the reaction was quenched by the addition of H2O (100 mL). The resulting mixture was transferred to 250-mL separatory funnel, and was washed with CH2Cl2 (2 x 100 mL). The combined organic layers were washed with a 0.5 M aqueous solution of HCl (4 x 50 mL), a solution of saturated aqueous NaHCO3 (50 mL), and a solution of saturated aqueous NaCl (50 mL). The organic layer was dried (Na2SO4), filtered, and the volatiles were removed in vacuo. The resulting viscous brown oil was purified by silica gel chromatography (99:1 CH2Cl2:MeOH) to provide the precursor to 3.23 as yellow oil (3.50 g, 19.3 mmol, 81% yield). Th

NOEt

O

precursor topiperidine 3.23

; the physical and spectral data for this compound were identical to those ported.64 1H NMR (400 MHz, CDCl3), 1:1 mixture of carbamate rotamers: m, 2H), 4.60−4.44 (m, 2H), 4.14 (q, J = 7.2 Hz, 2H), 1.80−1.30 (m, 6H), 2 Hz, 3H).

Syntheis of the azabicyclic precursor to 3.17 (shown on the left). This azbicycle was prepared by lithium aluminum hydride reduction of the carbamate in the precursor to 3.23; for a representative procedure, see azabicycle 3.6 (eq 3.9, see above). This azabicycle has been prepared

previously; the physical and spectral data for this compound were identical to those previously reported.74 1H NMR (400 MHz, CDCl3): δ 5.81 (s, 2H), 3.40 (s, 2H), 2.20 (s, 3H), 1.80−1.20 (m, 6H).

Synthesis of the azabicyclic precursor to 3.22 (the mesylate shown here on the left) was prepared by utilizing CbzCl, in place of ethylchloroformate, in the carbamate formation of the sequence shown in Scheme 3.7 (vide supra). A 25-mL round-bottom flask was charged with this mesylate

intermediate (0.220 mg, 0.840 mmol), collidine (4.20 mL), and 1,8-diazabicyclo[5.4.0]undec-7-ene (0.120 mL, 1.01 mmol, 1.2 equiv.), the flask was fitted

N

precursor topiperidine 3.17

N

OMs

OOBn

Me

(74) “Manipulation of substituents at nitrogen in tropanes, homotropanes, and dehydro- derivatives,”

−10914. Howarth, N. M.; Smith, C. R.; Malpass, J. R. Tetrahedron 1998, 54, 10899

Chapter 3, Page 111

with a reflux condenser, and the reaction mixture was allowed to stir at 160 oC for 24 h. At this time, the reaction mixture was cooled to 22 oC, and the reaction was quenched by the addition of was transferred combined organsaturated aqueoNaCl (15 mL).

2

H (2.4 mg, 0.0030 mmol, 0.05 equiv.) was dissolved in C6H6

n residue was purified by silica gel chromatography (9:1 CH2Cl2:MeOHIR (neat): 294δ 7.39−7.20 ((ddd, J = 17.2

CDCl3): δ 141.9, 137.1, 133.4, 130.7, 128.7, 127.6, 126.4, 115.7, 68.7, 67.6, 66.7, 43.0,

a solution of saturated aqueous NH4Cl (10 mL). The resulting mixture to a 50-mL separatory funnel, and washed with CH2Cl2 (4 x 10 mL). The ic layers were washed with a 0.5 M solution of HCl (3 x 15 mL), a us solution of NaHCO3 (20 mL), and a solution of saturated aqueous The organic layer was dried (Na2SO4), filtered, and the volatiles were

removed in vacuo. The resulting brown residue was purified by silica gel chromatography (gradient elution, 1:9 Et2O:hexanes followed by 1:1 Et2O:hexanes) to provide the precursor to 3.22 (shown here on the left) as colorless oil (0.140 g, 0.580 mmol, 70% yield).

IR (neat): 2939 (m), 2858 (w), 1704 (s), 1418 (s), 1306 (s), 1095 (s). 1H NMR (400 MHz, CDCl3), 1:1 mixture of carbamate rotamers: δ 7.39−7.28 (m, 5H), 6.09−6.01 (m, 2H), 5.17 (s, 2H), 4.60−4.56 (m, 2H), 1.82−1.63 (m, 3H), 1.47−1.43 (m, 1H), 1.35−1.26 (m, 2H). 13C NMR (100 MHz, CDCl3): δ 152.4, 137.0, 130.5, 130.1, 128.4, 127.9, 127.8, 66.4, 58.5, 58.5, 24.4, 23.5, 16.3. HRMS ES (m/z) Calcd for C15H17NO2 243.1254 (M)+, Found 243.1259.

Representative procedure for Mo-catalyzed asymmetric ring-opening/cross-metathesis reactions, synthesis of piperidine 3.7: In a N -filled glovebox, Mo complex

N

OOBn

OTBS

(0.250 mL) in a 4-mL vial. Styrene (68.0 µL, 0.593 mmol, 10 equiv.) was added to this solution by syringe. The resulting mixture was allowed to stir for 1 min, and added by syringe to a solution of azabicycle 3.6 (15.0 mg, 0.0592 mmol) in C6H6

(0.250 mL) in a 4-mL vial.75 The reaction mixture was allowed to stir for 1 h. At this time, the vial was removed from the glovebox, and the volatiles were removed in vacuo. The resulting dark brow

) to afford piperidine 3.7 as colorless oil (20.4 mg, 0.0570 mmol, 95%). 9 (s), 2930 (s), 2772 (s), 2363 (s), 1646 (w). 1H NMR (400 MHz, CDCl3): m, 5H), 6.48 (d, J = 15.6 Hz, 1H), 6.15 (dd, J = 15.6, 8.8 Hz, 1H), 5.79 , 10.4, 8.8 Hz, 1H), 5.15 (dd, J = 17.2, 1.6 Hz, 1H), 5.05 (dd, J = 10.4, 1.6

Hz, 1H), 3.77−3.68 (m, 1H), 2.77−2.60 (m, 1H), 2.58−2.48 (m, 1H), 2.18 (s, 3H), 1.87−1.78 (m, 2H), 1.60−1.51 (m, 2H), 0.87 (s, 9H), 0.05 (s, 6H). 13C NMR (100 MHz,

(75) Mo complex 6 is pre-treated with styrene to ensure formation of the chiral Mo benzylidene complex.

NMe

3.7

Chapter 3, Page 112

42.9, 41.5, 25.9, 18.3, −4.39. HRMS EI (m/z) Calcd for C22H35NOSi 357.2488 (M)+, Found 357.2482.

Representative procedure for desilylation of secondary alcohols in piperidine product, synthesis of piperidine 37-OH:76 A 4-mL vial was charged with 3.7 (20.4 mg,

0.0570 mmol), THF (0.500 mL), and TBAF (285 µL, 0.285 mmol, a 1.00 M solution in THF, 5 equiv.). The vial was tightly sealed with a Teflon cap, placed in a heating mantle at 65 oC, and the reaction mixture was allowed to stir for 1 h. At this time, the vial was cooled to 22 oC over 15 min, and the

revolatiles werechromatograp H as colorless oil (13.2

LC analysis in comparison with authentic racemic material, shown

moved in vacuo. The resulting yellow residue was purified by silica gel hy (4:1 CH2Cl2:MeOH) to afford piperidine 37-O

OH

mg, 0.0541 mmol, 95%). 1H NMR (400 MHz, CDCl3): δ 7.38−7.21 (m, 5H), 6.49 (d, J = 15.6 Hz, 1H), 6.16 (dd, J = 15.6, 8.8 Hz, 1H), 5.76 (ddd, J = 17.2, 10.4, 8.8 Hz, 1H), 5.17 (dd, J = 17.2, 1.6 Hz, 1H), 5.07 (dd, J = 10.4, 1.6 Hz, 1H), 3.82−3.70 (m, 1H), 2.77−2.60 (m, 1H), 2.58−2.48 (m, 1H), 2.20 (s, 3H), 2.00−1.92 (m, 2H), 1.55−1.43 (m, 3H). [α]D

20 −75.2 (c = 0.2, CHCl3) for a sample of 94% ee. The optical purity of this compound was determined by HPbelow: Chiralpak OD (4.6 x 250 mm), 99:1 hexanes:i-PrOH, 1.0 mL/min, λ = 254 nm, 94% ee.

1

3 −

−

−

NMe

OTBS

MeN

OTBS3.8

Piperidine 3.8 (homodimer byproduct). H NMR (400 MHz, CDCl ): δ 7.39 7.20 (m, 5H), 6.47 (d, J = 16.0 Hz, 1H), 6.13 (dd, J = 16.0, 8.8 Hz, 1H), 5.79 (ddd, J = 16.0, 10.0, 8.8 Hz, 1H), 5.52 5.47 (m, 2H), 5.13 (dd, J = 16.0, 1.6 Hz, 1H), 5.03 (dd, J = 10.0, 1.6 Hz, 1H), 3.75 3.62 (m, 2H),

(76) Determination of enantiomeric excess of 8, and in some other cases, involved the piperidine derived

NMe

3.7-OH

from TBAF deprotection of the secondary alcohol.

Chapter 3, Page 113

2.68−2.60 (m, 1H), 2.58−2.40 (m, 3H) 2.18−2.11 (m, 6H), 1.84−1.20 (m, 8H), 0.87 (s, 18H), 0.05 (s, 12H).

10.4 Hz, 1H), 3.75−3.67 1.86−1.74 (m, 2H), 1.62−

(m, 1H), 2.19 (s, 3H), 1.99−1.94 (m, 2H), 1.57−1.46 (m, 2H), 1.29−1.20 (br s, 1H). [α]D

20 −15.40 (c = 0.5, CHCl3) for a sample of 64% ee. The optical purity of this compound was determined by HPLC analysis in comparison with authentic racemic material, shown below: Chiralpak OD (4.6 x 250 mm), 98:2 hexanes:i-PrOH, 1.0 mL/min, λ = 254 nm, 64% ee.

Piperidine 3.9. IR (neat): 2949 (s), 2924 (s), 2779 (m), 1615 (w). 1H NMR (400 MHz, CDCl3): δ 7.55 (d, J = 8.0 Hz, 2H), 7.44 (d, J = 8.0 Hz, 2H), 6.49 (d, J = 15.6 Hz, 1H), 6.26 (dd, J = 8.4, 7.2 Hz, 1H), 5.76 (ddd, J = 18.0, 9.6, 9.6 Hz, 1H), 5.17 (d, J = 17.2 Hz, 1H), 5.06 (d, J =

(m, 1H), 2.69−2.61 (m, 1H), 2.54−2.47 (m, 1H), 2.17 (s, 3H), 1.48 (m, 2H), 0.87 (s, 9H), 0.06 (s, 6H). 13C NMR (100 MHz,

CDCl3): δ 141.7, 140.6, 136.2, 129.4, 126.5, 125.7, 125.6, 123.0, 115.9, 68.6, 67.5, 66.6, 42.9, 42.8, 41.5, 25.9, 18.3, −4.4. HRMS ES (m/z) Calcd for C23H35NOF3Si 426.2440 (M+H)+, Found 426.2451. Determination of enantiomeric excess of 3.9 involved the piperidine derived from desilylation of the silyl ether in 3.9.

Piperidine 9-OH. 1H NMR (400 MHz, CDCl3): δ 7.55 (d, J = 8.0 Hz, 2H), 7.45 (d, J = 8.0 Hz, 2H), 6.53 (d, J = 16.0 Hz, 1H), 6.26 (dd, J = 8.8, 8.0 Hz, 1H), 5.76 (ddd, J = 18.0, 9.6 Hz, 1H), 5.18 (d, J = 16.4 Hz, 1H), 5.09 (d, J = 10.0 Hz, 1H), 3.81−3.75 (m, 1H), 2.76−2.66 (m, 1H),

2.60−2.50

OTBS

CF3

NMe

3.9-OH

OH

CF3

NMe

3.9

2923 (s), 2854 (s), 2772 (m), 1508 (s). 1H NMR Piperidine 3.10. IR (neat): 2955 (s),

(400 MHz, CDCl3): δ 7.33 (d, J = 6.8 Hz, 2H), 6.84 (d, J = 8.8 Hz, 2H), 6.42 (d, J = 16.0 Hz, 1H), 6.00 (dd, J = 8.8, 7.2 Hz, 1H), 5.76 (ddd, J = 17.2, 10.0, 8.8 Hz, 1H), 5.18 (d, J

Chapter 3, Page 114

= 17.2 Hz, 1H), 5.07 (d, J = 10.0 Hz, 1H), 3.80 (s, 3H), OTBS3.80−3.67 (m, 1H), 2.63−2.55 (m, 1H), 2.55−2.46 (m,

68.7, 67.6, 66.8, 55.5, 43.1, 4C28H40NOSi 434.2879 (M+H)

5.17 (d, J = 17.2 Hz, 1H),

1H), 2.70−2.60 (m, 1H), 2.60−(m, 3H). The optical purity comparison with authentic racmm), 99:1 hexanes:i-PrOH, 1.0

1H), 2.18 (s, 3H), 1.86−1.76 (m, 2H), 1.61−1.48 (m, 2H), 0.85 (s, 9H), 0.04 (s, 6H). 13C NMR (100 MHz, CDCl3): δ 159.3, 141.9, 131.1, 130.2, 129.9, 127.6, 115.6, 114.2, 2.9, 41.4, 26.0, 18.3, −4.4. HRMS ES (m/z) Calcd for

+, Found 434.2889. [α]D20 −31.99 (c = 0.5, CHCl3) for a

sample of 89% ee. Determination of enantiomeric excess of 3.10 involved the piperidine derived from desylilation of the secondary silyl ether in 3.10.

Piperidine 3.10-OH. 1H NMR (400 MHz, CDCl3): δ 7.32−7.29 (m, 2H), 6.85 (d, J = 8.8 Hz, 2H), 6.43 (d, J = 15.6 Hz, 1H), 6.01 (dd, J = 15.2, 8.4 Hz, 1H), 5.78 (ddd, J = 17.2, 10.0, 8.8 Hz, 1H), Me

OH

NMe

OMe3.10

5.07 (d, J = 10.0 Hz, 1H), 3.80 (s, 3H), 3.80−3.70 (m, 2.50 (m, 1H), 2.22 (s, 3H), 2.02−1.90 (m, 2H), 1.60−1.44 of this compound was determined by HPLC analysis in emic material, shown below: Chiralpak OD (4.6 x 250 mL/min, λ = 254 nm, 89% ee.

Piperidine 3.13. IR (neat): 2949 (s), 2930 (s), 2854 (s), 2364 (m), 2338 (m). 1H NMR (400 MHz, CDCl3): δ 7.46−7.40 (m, 1H), 7.18−7.12 (m, 3H), 6.81 (d, J = 16.0 Hz, 1H), 6.03 (dd, J = 8.8, 7.2 Hz, 1H), 5.77 (ddd, J = 18.4, 10.0, 10.0 Hz, 1H), 5.15 (d, J = 16.8 Hz, 1H), 5.04 (d, J = 10.4 Hz, 1H), 3.73 (dddd, J =

15.2, 10.4, 4.8 Hz, 1H), 2.68−2.62 (m, 1H), 2.53−2.48 (m, 1H), 2.35 (s, 3H), 2.19 (s, 3H), 1.86−1.76 (m, 2H), 1.63−1.49 (m, 2H), 0.88 (s, 9H), 0.07 (s, 6H). C NMR (100 MHz, CDCl ): δ 141.9, 136.1, 135.3, 134.7, 130.4, 128.5, 127.5, 126.3, 125.8, 115.7, 68.7,

N

OMe3.10-OH

3.13

NMe

Me

OTBS

13

3

Chapter 3, Page 115

67.6, 67.0, 43.2, 43.0, 41.5, C23H38NOSi 372.2723 (M+H)+

excess of 3.13 involved the pipePiperi(m, 1

was determined by HPLC analysis in comparison with autx 250 mm), 99.8:0.2 hexanes:i-P

25.9, 20.0, 18.3, −4.4. HRMS ES (m/z) Calcd for , Found 372.2728. Determination of the enantiomeric ridine derived from desilylation of the silyl ether in 3.13.

dine 3.13-OH 1H NMR (400 MHz, CDCl3): δ 7.45−7.40 H), 7.18−7.13 (m, 3H), 6.73 (d, J = 15.6 Hz, 1H),

6.11−6.04 (m, 1H), 5.88−5.76 (m, 1H), 5.20 (d, J = 17.2 Hz, 1H), 5.10 (d, J = 9.6 Hz, 1H), 3.85−3.75 (m, 1H), 2.80−2.55 (m, 2H), 2.34 (s, 3H), 2.26 (s, 3H), 2.04−1.95 (m, 2H), 1.70−1.50

(m, 2H), 1.22 (br s, 1H). The optical purity of this compound

3.13-OH

NMe

hentic racemic material, shown below: Chiralpak OJ (4.6 rOH, 1 mL/min, λ = 254 nm, >98% ee.

Piperidine 3.14. IR (neat): 2949 (s),

2924 (s), 2855 (s), 2363 (m), 1463 (w). 1H NMR (400 MHz, CDCl3): δ 7.54−7.40 (m, 2H), 7.26−7.22 (m, 1H), 7.09−7.05 (m, 1H), 6.83 (d, J = 15.6 Hz, 1H), 6.09 (dd, J = 8.8, 6.8 Hz, 1H), 5.77 (ddd, J = 18.4, 9.6 Hz, 1H), 5.15 (d, J = 17.2 Hz, 1H), 5.00 (d, J = 10.0 Hz, 1H),

3.75−3.68 (m, 1H), 2.73−2.67 (m, 1H), 2.53−2.48 (m, 1H), 2.19 (s, 3H), 1.85−1.76 (m, 2H), 1.60−1.47 (m, 2H), 0.87 (s, 9H), 0.04 (s, 6H). 13C NMR (100 MHz, CDCl3): δ 141.8, 136.9, 136.3, 132.1, 129.5, 128.9, 127.7, 127.1, 123.6, 115.8, 68.6, 67.5, 66.6, 43.0, 42.9, 41.5, 25.9, 18.3, −4.4. HRMS ES (m/z) Calcd for C22H35NOSiBr 436.1671 (M+H) +, Found 436.1678. [α] 20 −30.89 (c = 0.5, CHCl ) for a sample of 88% ee

it

1H), 2.23 (s, 3H), 2.03−1.93 (m, 2H), 1.60−1.52 (m, 2H), 1.40 (br s, 1H). The optical

OH

Me

3.14

NMe

OTBS

Br

D 3

omeric excess of 3.14 involved the piperidine derived from her in 3.14.

Piperidine 3.14-OH 1H NMR (400 MHz, CDCl3): δ 7.57−7.50 (m, 2H), 7.30−7.24 (m, 1H), 7.11−7.07 (m, 1Η), 6.85 (d, J = 15.6 Hz, 1H), 6.12 (dd, J = 8.8, 6.8 Hz, 1H), 5.77 (ddd, J = 18.4, 9.6 Hz, 1H), 5.18 (d, J = 17.2 Hz, 1H), 5.09 (d, J = 10.0 Hz, 1H), 3.83−3.70 (m, 1H), 2.80−2.70 (m, 1H), 2.62−2.55 (m, OH

NMe

. Determination of enantdesilylation of the silyl e

OH

Br

3.14-

Chapter 3, Page 116

purity of this compound was determined by HPLC analysis in comparison with authentic racemic material, shown below: Chiralcel OJ (4.6 x 250 mm), 99.8:0.2 hexanes:i-PrOH, 1.0 mL/min, λ = 254 nm, 88% ee.

Piperidine 3.15. IR (neat): 3030 (m), 2943 (s), 2924 (s), 2773 (s), 1451 (s). 1H NMR (400 MHz, CDCl3): δ 7.30−7.10 (m, 10H), 6.41 (d, J = 16.0 Hz, 1H), 6.10 (dd, J = 8.8, 7.2 Hz, 1H), 5.72 (ddd, J = 18.8, 10.4, 8.8 Hz, 1H), 5.10 (dd, J = 17.2, 1.6 Hz, 1H), 5.00 (dd, J = 10.0, 1.6 Hz, 1H), 4.48 (s, 2H),

3.48−3.38 (m, 1H), 2.59−2.53 (m, 1H), 2.47−2.41 (m, 1H), 2.12 (s, 3H), 2.04−1.92 (m, 2H), 1.57−1.44 (m, 2H). 13C NMR (100 MHz, CDCl3): δ 142.9, 139.8, 137.6, 134.3130.6, 128.9, 128.6, 127.39.9. HRMS EI (m/z) C−14.80 (c = 0.1, CHCl3) fdetermined by HPLC anbelow: Chiralpak OD (4

, 7, 127.6, 127.5, 126.7, 114.9, 74.5, 69.6, 67.5, 66.6, 41.6, 40.0, alcd for C23H28NO 334.2171 (M+H)+, Found 334.2174. [α]D

20 or a sample of 96% ee. The optical purity of this compound was alysis in comparison with authentic racemic material, shown

.6 x 250 mm), 98:2 hexanes:i-PrOH, 1.0 mL/min, λ = 254 nm, 96% ee.

NMe

OBn

3.15

Chapter 3, Page 117

Piperidine 3.16. IR (neat): 2961 (s), 2930 (s), 2854 (m), 1256 (m). 1H NMR (400 MHz, CDCl3): δ 7.40−7.20 (m, 5H), 6.48 (d, J = 16.0 Hz, 1H), 6.15 (dd, J = 9.7, 7.2 Hz, 1H), 5.74 (ddd, J = 18.4, 9.6

NMe

OTBS

, 9.6 Hz, 1H), 5.15 (d, J = 16.8 Hz, 1H), 5.03 (d, J = 10.0 Hz, 1H), 4.07 (br s, 1H), 3.10−3.00 (m, 1H), 2.99−2.90 (m,

1H), 2.20 (s, 3H), 1.80−1.55 (m, 4H), 0.91 (s, 9H), 0.04 (s, 6H). 13C NMR (100 MHz, CDCl3): δ 142.0, 137.2, 133.7, 130.8, 128.7, 127.5, 126.4, 115.9, 64.9, 63.0, 62.2, 41.8, 40.9, 26.0, 18.3, −4.67, −4.70. HRMS EI (m/z) Calcd for C22H36NOSi 358.2566 (M+H)+, Found 358.2571. Determination of enantiomeric excess of 3.16 involved the piperidine derived from desilylation of the silyl ether in 3.16.

Piperidine 3.16-OH 1H NMR (400 MHz, CDCl3, TMS): δ 7.38−7.20 (m, 5H), 6.49 (d, J = 16.0 Hz, 1H), 6.12 (dd, J = 16.0, 8.8 Hz, 1H), 5.72 (ddd, J = 18.4, 9.6, 9.6 Hz, 1H), 5.20 (d

(r

material, shown below: C

3.16

OH

, J = 17.2 Hz, 1H), 5.06 (d, J = 16 Hz, 1H), 4.17 (s, 1H), 3.10−3.00 (m, 1H), 2.98−2.90 (m, 1H), 2.25 (s, 3H), 1.80−1.70 c = 0.01, CHCl3) for a sample of 84% ee. The optical purity of mined by HPLC analysis in comparison with authentic racemic hiralpak OJ (4.6 x 250 mm), 97:3 hexanes:i-PrOH, 1.2 mL/min,

λ = 254 nm, 84% ee.

(m, 4H). [α]D20 −79.98

this compound was dete

NMe

3.16-OH

Piperidine 3.17. IR (neat): 2930 (s), 2848 (m) 2773 (s), 1445 (w). 1H NMR (400 MHz, CDCl3): δ 7.40−7.20 (m, 5H), 6.49 (d, J = 16.0 Hz, 1H), 6.18 (dd, J = 8.8, 7.2 Hz, 1H), 5.82−5.73 (m, 1H), 5.15 (dd, J = 17.6, 2.0 Hz, 1H), 5.03 (dd, J = 10.4, 1.6

Hz, 1H), 2.60−2.54 (m, 1H), 2.47−2.42 (m, 1H), 2.21 (s, 3H), 1.80−1.42 (m, 6H). 13C NMR (100 MHz, CDCl3): δ 142.7, 137.4, 134.3, 130.5, 128.8, 127.5, 126.4, 115.5, 69.4, 68.5, 42.3, 33.9, 33.8, 24.0. HRMS EI (m/z) Calcd for C16H21N 227.1674 (M)+, Found

3.17

NMe

Chapter 3, Page 118

227.1679. [α]D20 −31.19 (c = 0.5, CHCl3) for a sample of 80% ee. The optical purity of

r:%

mined by HPLC analysis in comparison with authentic racemic Chiralpak OD (4.6 x 250 mm), 98:2 hexanes:i-PrOH, 1.0

ee.

this compound was detematerial, shown belowmL/min, λ = 254 nm, 80

Piperidine 3.19. IR (neat): 2955 (m), 2923 (m), 2855 (w), 1697 (s), 1073 (s). 1H NMR (400 MHz, CDCl3): δ 7.37−7.19 (m, 10H), 6.68 (dd, J = 16.0, 8.4 Hz, 1H), 6.40 (d, J = 15.2 Hz, 1H), 6.28 (ddd, J = 17.2, 10.0, 7.2 Hz, 1H), 5.19−5.01 (m, 4H), 4.98−4.92 (m, 1H), 4.86−4.78 (m, 1H), 4.16 (dddd, J = 7.6, 4.4,

4.4, Hz 1H), 2.10−1.84 13

NCbz3.19

OTBS

(m, 4H), 0.89 (s, 9H), 0.07 (s, 3H), 0.03 (s, 3H). C NMR (100 MHz, CDCl3): δ 156.1, 141.5, 137.4, 136.9, 132.8, 130.5, 128.6, 128.5, 128.2, 128.0, 127.3, 126.5, 114.9, 67.4, 64.9, 52.4, 52.3, 36.7, 36.5, 26.0, 18.2, −4.7. HRMS EI (m/z) Calcd for C29H39NO3Si 477.2693 (M)+, Found 477.2699. [α]D

20 92.8 (c = 0.08, CHCl3) for a sample of 90% ee. The optical purity of this compound was determined by HPLC analysis in comparison with authentic racemic material, shown below: Chiralpak OD (4.6 x 250 mm), 99.5:0.5 hexanes:i-PrOH, 1.0 mL/min, λ = 254 nm, 90% ee.

Chapter 3, Page 119

Piperidine 3.21. IR (neat): 2955 (s), 2924 (s), 2861 (m), 1697 (s), 1400 (s). 1H NMR (400 MHz, CDCl3): δ 7.37−7.22 (m, 10H), 6.47 (d, J = 16.0 Hz, 1H), 6.24 (dd, J = 16.0, 6.0 Hz, 1H), 5.93 (ddd, J = 16.4, 10.8, 6.0 Hz, 1H), 5.21−5.10 (m, 5H), 5.00−

NCbz

OTBS

3.21 4.98 (m, 1H), 4.17 (dddd, J = 11.2, 7.2, 7.2, 4.0 Hz, 1H), 2.10−2.05 (m, 2H), 1.83−1.67 (m, 2H), 0.90 (s, 9H), 0.09 (s, 6H). 13C NMR (100 MHz, CDCl3): δ 155.8, 139.7, 137.0, 136.8, 130.8, 130.6, 128.7, 128.6, 128.1, 128.0, 127.7, 126.5, 115.5, 67.5, 62.3, 53.6, 53.5, 38.5, 38.0, 26.0, 18.3, −4.37. HRMS EI (m/z) Calcd for C29H39NO3Si 477.2699 (M)+, Found 477.2693. [α]D

20 39.9 (c = 0.01, CHCl3) for a sample of 85% ee. The optical purity of this compound was determined by HPLC analysis in comparison with authentic racemic material, shown below: Chiralpak OD (4.6 x 250 mm), 99.5:0.5 hexanes:i-PrOH, 1.0 mL/min, λ = 254 nm, 85% ee.

Piperidine 3.22. IR (neat): 2936 (w), 1697 (s), 1407 (m). 1H NMR (400 MHz, CDCl ): δ 7.39−7.21 (m, 10H), 6.47 (d, J = 16.0 Hz, 1H), 6.34 (dd, J = 16.0, 7.1, Hz, 1H), 5.98 (ddd, J

3

3 , 137.2, 136.9, 130.6, 128.6, 128.5, 128.1, 128.0, 127.5, 126.4, 115.4, 67.3, 53.5, 52.5, 52.4, 29.1, 28.3, 15.1. HRMS EI (m/z) Calcd for C23H25NO2 347.1885 (M)+, Found 347.1892. [α]D

20 53.58 (c 0.05, CHCl3) for a sample of 82% ee. The optical purity of this compound was determined by HPLC analysis in comparison with authentic racemic material, shown below: Chiralpak OD (4.6 x 250 mm), 98:2 hexanes:i-PrOH, 1.0 mL/min, λ = 254 nm, 82% ee.

= 17.2, 10.5, 5.8 Hz, 1H), 5.20 (d, J = 12.3 Hz, 1H), 5.14 (d, J =

12.3 Hz, 1H), 5.14−5.10 (m, 2H), 5.05−4.97 (m, 1H), 4.88−4.85 (m, 1H), 1.94−1.72 (m, 5H), 1.60−1.54 (m, 1H). 13C NMR (100 MHz, CDCl ): δ 156.0, 139.6

NCbz

3.22

Chapter 3, Page 120

Piperidine 3.23 IR (neat): 2936 (w), 2357 (m), 2332 (m), 1690 (s). H NMR (400 MHz, CDCl ): δ 7.34−7.18 (m, 5H), 6.48 (d, J = 16.4 Hz, 1H), 6.31 (dd, J = 16.4, 9.2 Hz, 1H), 5.95 (ddd, J = 17.6, 10.8, 6.0 Hz, 1H), 5.14 (dd, J = 17.6, 1.6 Hz, 1H), 5.11

(dd, J = 10.8, 1.6 Hz, 1H), 4.97−4.92 (m, 1H), 4.82−4.77 (m, 1H), 4.15 (q, J = 7.2 Hz, 2H), 1.92−1.62 (m, 4H), 1.58 (br, 2H), 1.25 (t, J = 7.2 Hz, 3H). 3C NMR (100 MHz, CDCl ): δ 156.3, 139.8, 137.4, 131.0, 130.4, 128.6, 127.5, 126.4, 115.3, 61.4, 52.4, 52.2, 29.0, 28.3, 15.2, 14.8. HRMS EI (m/z) mass Calcd for C1 NO 285.1728 (M) , Found 285.1720. [α] 52.87 (c = 1.0, CHCl ) for a sample of 90% ee. The optical purity of this compound was determined by HPLC analysis in comparison with authentic racemic material, shown below: Chiralpak OD (4.6 x 250 mm), 98:2 hexanes:i-PrOH, 1.0 mL/min, λ = 254 nm, 90% ee.

. 1

3

1

3

8H23 2+

D20

3

NCO2Et3.23

Piperidine 3.24. IR (neat): 2955 (m), 2930 (m), 2855 (w), 1697 (s), 1073 (s). 1H NMR (400 MHz, CDCl3): δ 7.34−7.19 (m, 5H), 6.65 (dd, J = 15.6 8.4 Hz, 1H), 6.44 (d, J = 16.0 Hz, 1H), 6.22 (ddd, J = 17.6, 10.4, 7.6 Hz, 1H), 5.10 (dd, J = 17.2, 1.6 Hz, 1H), 5.03 (dd, J = 10.0, 1.6 Hz, 1H), 4.96−4.90 (m, 1H), 3.24

NCO2Et

OTBS

Chapter 3, Page 121

4.82−4.76 (m, 1H), 4.20−4.09 (m, 2H), 2.02−1.84 (m, 4H), 1.26 (t, J = 7.2 Hz, 3H), 0.89 (s, 9H), 0.07 (s, 3H), 0.04 (s, 3H). 13C NMR (100 MHz, CDCl3): δ 156.3, 141.6, 137.5, 133.1, 130.3, 128.5, 127.3, 126.5, 114.7, 64.9, 61.4, 52.2, 52.1, 36.7, 36.5, 26.0, 18.2, 14.8, −4.8. HRMS EI (m/z) Calcd for C24H37NO3Si 415.2543 (M)+, Found 415.2551. Determination of enantiomeric excess of 3.24 involved the piperidine derived from desilylation of the silyl ether in 3.24.

OH Piperidine 3.24-OH 1H NMR (400 MHz, CDCl3): δ 7.36−7.21 (m, 5H), 6.54−6.48 (m, 2H), 6.20 (ddd, J = 16.8, 10.0, 6.4 Hz, 1H), 5.18 (dd, J = 17.2, 1.2 Hz, 1H), 5.10 (dd, J = 10.0, 1.2 Hz, 1H), 5.00−4.94 (m, 1H), 4.86−4.80 (m, 1H), 4.24−4.20 (m, 2H), 4.16 (q, J = 7.2 Hz, 2H), 2.12−1.96 (m, 4H), 1.26 (t, J = 7.2 Hz,

3H). [α]D20 27.0 (c = 0.1, CHCl3) for a sample of 79% ee. The optical purity of this

compound was determined by HPLC analysis in comparison with authentic racemic material, shown below: Chiralpak OD (4.6 x 250 mm), 90:10 hexanes:i-PrOH, 1.0

L/min, λ = 254 nm, 79% ee.

3.24-OH

NCO2Et

m

Piperidine 3.25. IR (neat): 2949 (s), 2924 (s), 2855 (m), 1697 (s), 1508 (s), 1250 (s). 1H NMR (400 MHz, CDCl3, TMS): δ 7.38−7.30 (m, 7H), 7.22 (d, J = 8.8 Hz, 2H), N

CbzOMe3.25

6.81 (d, J = 8.8 Hz, 2H), 6.53 (dd, J = 16.0, 8.4 Hz, 1H), 6.35 (d, J = 16.0 Hz, 1H), 6.28 (ddd, J = 17.2, 10.4, 7.6

Hz, 1H), 5.19−5.08 (m, 2H), 5.07−5.06 (m, 1H), 5.01 (dd, J = 12.0, 1.2 Hz, 1H), 4.94−4.90 (m, 1H), 4.84−4.80 (m, 1H) 4.18−4.11 (m, 1H), 3.80 (s, 2H), 2.05−1.85 (m, 4H), 0.89 (s, 9H), 0.07 (s, 3H), 0.05 (s, 3H). 13C NMR (100 MHz, CDCl3): δ 200.8, 189.7, 158.8, 141.2, 130.4, 129.7, 128.2, 127.8, 127.7, 127.4, 114.5, 113.6, 67.0, 64.7, 55.1, 52.2, 36.5, 36.3, 25.7, 17.9, −0.2, −5.1. HRMS ES (m/z) Calcd for C30H42NO4Si 508.2883 (M+H)+, Found 508.2871. [α]D

20 23.99 (c = 0.1, CHCl3) for a sample of 80%

OTBS

Chapter 3, Page 122

ee. The optical purity of this compound was determined by HPLC analysis in mparison with authentic racemic material, shown below: Chiralpak OD-R (4.6 x 250

-co

PrOH, 1.0 mL/min, λ = 254 nm, 80% ee. mm), 99.5:0.5 hexanes:i

Piperidine 3.26. IR (neat): 2949 (s), 2924 (s), 2855 (m), 1697 (s), 1400 (m), 1073 (s). H NMR (400 MHz, CDCl , TMS): δ 7.38−7.26 (m, 7H), 7.20−7.10 (m, 2H), 6.63 (d, J = 16.0 Hz, 1H), 6.55 (dd, J = 16.0, 8.0 Hz, 1H), 6.28 (ddd, J = 17.2, 10.0, 7.2 Hz, 1H), 5.15 (s, 2H), 5.10 (dd, J = 17.2, 1.2 Hz, 1H), 5.02

(dd, J = 10.4, 1.2 Hz, 1H), 5.00−4.96 (m, 1H), 4.84−4.80 (m, 1H), 4.16 (dd, J = 8.0, 4.4 Hz, 1H), 2.22 (s, 3H), 2.05−1.87, (m, 4H) 0.88 (s, 9H), 0.03 (s, 3H), 0.02 (s, 3H). C NMR (100 MHz, CDCl ): δ 156.1, 141.5, 137.0, 136.4, 135.5, 134.0, 130.2, 128.6, 128.2, 128.1, 128.0, 127.2, 126.0, 125.6, 114.9, 67.4, 65.0, 52.6, 52.4, 36.8, 36.6, 26.0, 19.9, 18.2, −4.8. HRMS ES (m/z) Calcd for C30H42NO3Si 492.2934 (M+H)+, Found 492.2928. [α]D

20 47.99 (c = 0.3, CHCl3) for a sample of 86% ee. The optical purity of this compound was determined by HPLC analysis in comparison with authentic racemic material, shown below: Chiralpak OD-R (4.6 x 250 mm), 99.5:0.5 hexanes:i-PrOH, 1.0 mL/min, λ = 254 nm, 86% ee.

NCbz

OTBS

3.26

Me1

3

13

3

Chapter 3, Page 123

Piperidine 3.30. IR (neat): 2924 (s), 2848 (s), 2766 (w), 1099 (m). 1H NMR (400 MHz, CDCl3): δ 5.73 (ddd, J = 18.8,10.0, 10.0 Hz, 1H), 5.50 (dd, J = 15.2, 6.4 Hz, 1H), 5.28 (dd, J = 15.2, 8.8 Hz, 1H), 5.12 (dd, J = 17.6, 1.2 Hz, 1H), 5.02 (dd, J = 10.0, 1.2 Hz, 1H), 3.65 (dddd, J = 15.2, 10.4, 4.4, 4.4 Hz, 1H),

2.43 (dd, J = 8.8, 8.8 Hz, 1H), 2.36 (dd, J = 8.8, 8.8 Hz, 1H), 2.11 (s, 3H), 1.98−1.86 (m, 1H), 1.80−1.58 (m, 6H), 1.54−1.38 (m, 2H), 1.30−1.00 (m, 6H), 0.87 (s, 9H), 0.05 (s, 6H). 13C NMR (100 MHz, CDCl3): δ 124.2, 138.0, 130.8, 115.4, 68.8, 67.6, 66.6, 43.3, 43.1, 41.1, 40.3, 33.0, 26.4, 26.2, 26.0, 18.3, −4.4. HRMS ES (m/z) Calcd for C22H42NOSi 364.3036 (M+H) +, Found 364.3035. Determination of enantiom

NMe

3.30

OTBS

eric excess of 3.30 involved the piperidine derived from desilylation of the silyl ether in 30.

Piperidine 3.30-OH 1H NMR (400 MHz, CDCl3): δ 5.76 (ddd, J = 18.8, 9.2, 9.2 Hz, 1H), 5.51 (dd, J = 15.6, 6.4 Hz, 1H), 5.31 (dd, J = 15.2, 7.6 Hz, 1H), 5.15 (dd, J = 17.2, 1.6 Hz, 1H), 5.05 (dd, J = 10.4, 1.6 Hz, 1H), 3.75−3.67 (m, 1H), 2.52−2.40 (m, 2H), 2.14 (s, 3H), 1.98−1.86 (m, 2H), 1.72−1.60 (m, 5H),

1.51−1.36 (m, 2H), 1.25−0.83 (m, 6H). The optical purity of this compound was determined by GC analysis in comparison with authentic racemic material, shown below: β-dex chiral column, 130 oC, 20 psi.

NMe

3.30-OH

OH

Chapter 3, Page 124

Chapter 3, Page 125

Chapter 3, Page 126

Chapter 3, Page 127

Chapter 3, Page 128

Chapter 3, Page 129

Chapter 3, Page 130

Chapter 3, Page 131

Chapter 3, Page 132

Chapter 3, Page 133

Chapter 3, Page 134

Chapter 3, Page 135

Chapter 3, Page 136

Chapter 3, Page 137

Chapter 3, Page 138

Chapter 3, Page 139

Chapter 3, Page 140

Chapter 3, Page 141

Chapter 3, Page 142

Chapter 3, Page 143

Chapter 3, Page 144

Chapter 3, Page 145

Chapter 3, Page 146

Chapter 3, Page 147

Chapter 3, Page 148

Chapter 3, Page 149

Chapter 3, Page 150

Chapter 3, Page 151

Chapter 3, Page 152

Chapter 3, Page 153

Chapter 3, Page 154

Chapter 3, Page 155

Chapter 3, Page 156

Chapter 3, Page 157

Chapter 3, Page 158

Chapter 3, Page 159

Chapter 3, Page 160

Chapter 3, Page 161

Chapter 3, Page 162

Chapter 3, Page 163

Chapter 3, Page 164

Chapter 3, Page 165

Chapter 3, Page 166

Chapter 3, Page 167

Chapter 3, Page 168

Chapter 3, Page 169

Chapter 3, Page 170

Chapter 3, Page 171

Chapter 3, Page 172

Chapter 3, Page 173

Chapter 3, Page 174

Chapter 3, Page 175

Chapter 3, Page 176

Chapter 3, Page 177

Chapter 3, Page 178

Chapter 3, Page 179

Chapter 3, Page 180

Chapter 3, Page 181

Chapter 3, Page 182

Chapter 3, Page 183

Chapter 3, Page 184

Chapter 3, Page 185

Chapter 3, Page 186

Chapter 3, Page 187

Chapter 3, Page 188

Chapter 4, Page 189

Chapter Four

FUNCTIONALIZATION OF PIPERIDINE PRODUCTS

4.1 Introduction

In the preceding chapter several biologically active alkaloids were presented and discussed as possible targets for natural product synthesis utilizing our newly developed method for the catalytic enantioselective synthesis of piperidines. To demonstrate the utility of such enantioenriched piperidines we undertook studies directed at the functionalization of these products. Our initial goal was to elaborate AROM/CM piperidine products to intermediates that could potentially be applied to the synthesis of the alkaloids lobeline and sedinone77 (Scheme 4.1). Of special interest to us was that the only structural difference between these two natural products is that lobeline possess a phenyl ketone while sedinone has a methyl ketone. Thus we envisaged an approach where an advanced intermediate could be diverted in one step to either of the two natural products.

NCO2Et

NCO2Et

OHO

NMe

MeOAlkylation with

PhMgCl or MeMgCl

NCO2Et

O

NMe

MeO

4.4 90% ee with Mo-catalyst

NCO2Et

OO

NMe

MeODiastereoselective

Reduction

NMe

R

O OH

R = Ph, lobelineR = Me, sedinone

Rh-catalyzedHydroboration

Selective OlefinFunctionalization

Scheme 4.1

4.1

4.2

4.3

(77) “Alkaloids of Some Asian Sedum Species,” Kim, J. H.; Hart, H. T.; Stevens, J. F. Phytochemistry, 1996, 41, 1319−1324.

Chapter 4, Page 190

4.1a Retrosynthetic Analysis of Alkaloid Natural Products. Our initial retrosynthetic analysis of lobeline and sedinone is shown in Scheme 4.1 (see above). We expected to arri inreb amide in 4.1. Our plan was to convert the carbamate in 4.1 to an N-Me group in the same reaction

essel as the alkylation of the Weinreb amide, given the stability of tetrahedral ed from Weinreb amide alkylations.78 We projected that 4.1 would be

accessi

that the terminal olefin moiety in 4.4 could be differentiated from the more substituted styrenyl olefin and converted to the Weinreb amide in 4.3.

4.2 Initial Approach to Piperidine Alkaloids

4.2a Synthesis of Weinreb Amide Intermediate. The first objective of preparing Weinreb amide 4.3 was met readily and the sequence to arrive at this intermediate is shown in Scheme 4.2 (see below). To this end, treatment of piperidine 4.4 with [3.3.1]-9-borabicyclononane followed by an oxidative work-up delivers primary alcohol 4.5 in 97% yield. A series of three well-known reactions were then employed to access 4.3. To this result, oxidation of the primary alcohol in 4.5 to the aldehyde with pyridinium chlorochromate, followed by further oxidation to the carboxylic acid with sodium chlorite, and ultimately peptide coupling with N,O-dimethylhydroxyl amine hydrochloride delivers piperidine 4.3 in 78% yield (over three steps). With this

ve at either natural product from the alkylation of the We

vintermediates form

ble from benzylic ketone 4.2. No problems were foreseen in performing a diastereoselective reduction of the benzylic alcohol in 4.2 since stereoselective reduction of benzylic ketones has been well documented in the literature.79 We anticipated, and found it to be true, that the most difficult functionalization in this synthetic sequence would be conversion of the styrenyl olefin in 4.3 to the benzylic ketone in 4.2. A known method for the benzylic oxidation of styrenyl olefins is through Rh-catalyzed hydroboration/oxidation reactions.80 These reports, however, generally relate to the hydroboration reactions of terminal styrenes. We did not know whether these methods would be applicable to a substituted styrene such as the one in 4.3. The starting point in this synthesis was to be piperidine 4.4, accessible in 90% ee through Mo catalysis. We speculated

(78) “N-methoxy-n-methylamides as Effective Acylating Agents,” Nahm, S.; Weinreb, S. M. Tetrahedron Lett. 1981, 22, 3815−3818. (79) (a) “Asymmetric Boron-catalyzed Reactions,” Deloux, L.; Srebnik, M. Chem. Rev. 1993, 93, 763−784. (b) “New Chiral Phosphorus Ligands for Enantioselective Hydrogenation ,” Tang, W.; Zhang, X. Chem. Rev. 2003, 103, 3029−3070. (80) “On the Origin of Regio- and Stereoselectivity in the Rhodium-Catalyzed Vinylarenes Hydroboration Reaction,” Daura-Oller, E.; Segarra, A. M.; Poblet, J. M.; Claver, C.; Fernandez, E.; Bo, C. J. Org. Chem. 2004, 69, 2669−2680.

Chapter 4, Page 191

intermediate in hand we began to study Rh-catalyzed hydroboration/oxidation reactions of the styrenyl olefin moiety.

CO2EtH2O2, NaOH (aq),

1 h CO2Et

4.5 97%

OH

i) PCC, NaOAc,4 Å MS, CH2Cl2, 2 h

N2

4.3 78%over 3 steps

O

NMeO

Meii) NaClO2, NaHPO4,t-BuOH, H2O, tetra-

hiii) HN(OMe)MeHCl, EDC,HOBt, NEt3, CH2Cl2, 2 h

4.4

Scheme 4.2

4.2b Hydroboration Attemps of Styrenyl Olefin. Our attempts to perform Rh-catalyzed hydroboration/oxidation reactions of substituted styrenyl olefins are shown in Scheme 4.3 (see below). It is well known that Wilkinson’s catalyst can perform the regioselective hydroboration/oxidation of styrenyl olefins to deliver b

N

9-BBN, THF;

N

CO Et

methylethylene,1

enzylic alcohols.81 e found that Wilkinson’s catalyst, in the

oxidation of the styrenyl olefin in 4.3.

Unfortunately, under a variety of conditions, wpresence of catechol borane, was unreactive toward the styrenyl olefin in 4.3. To determine whether the Weinreb amide in 4.3 was responsible for this lack of reactivity we studied the hydroboration/oxidation reaction with piperidine 4.4, hoping to perform a double oxidation. Based on steric considerations, we expected the terminal olefin to be converted to a primary alcohol that could then be converted to a Weinreb amide. As expected, Wilkinson’s catalyst cleanly converted the terminal olefin to a primary alcohol. Dauntingly, however, we found that Wilkinson’s catalyst was again unreactive toward the styrenyl olefin moiety of the piperidine and found the major product from the reaction to be 4.5. After these disappointing results, we pursued an alternative approach to the

(81) “Transition-metal promoted hydroborations of alkenes, emerging methodology for organic transformations,” Burgess, K.; Ohlmeyer, M. J. Chem. Rev. 1991, 91, 1179−1191.

Chapter 4, Page 192

NCO2Et

1 mol % Rh(PPh3)3Cl

OBH

O

H2O2, NaOHTHF,22 oC

NCO2Et

OH

4.5 50%4.4

1 mol % Rh(PPh3)3Cl

OBH

O

variousconditions

4.3

<2% conv.NCO2Et

O

NMeO

Me

Scheme 4.3

4.2c Attempts at Wacker Oxidation of a Substituted Styrenyl Olefin. Around this time we came across an earlier report in the literature, in which a tetrahydropyran with a

-

styrenyl substituent at the 2-position underwent a Wacker oxidation to deliver a benzylic ketone.82 The researchers reporting this reaction proposed that the oxygen directed Pd to the proximal carbon of the styrene, resulting in the formation of a benzylic ketone as the major product. We were hopeful this directing effect would also apply to carbamateprotected piperidines and studied the Wacker oxidation of 4.3. To our misfortune, the attempts proved unfruitful (eq. 4.1). The reaction conditions screened involved the use of either palladium (II) chloride or palladium (II) acetate as catalytic palladium sources and the use of either copper (II) acetate monohydrate or copper (I) chloride as re-oxidant. A variety of solvent mixtures and temperatures were also screened; in all cases the starting piperidine 4.3 was recovered in >98% yield.

variousconditions

4.3

<2% conv.NCO2Et

O

NMeO

Me

PdCl2 or Pd(OAc)2H2O-Cu(OAc)2 or CuCl

O2 (4.1)

4.2d Attempts at Performing Benzylic Oxidations of Alkyl-arenes . At this point,

we decided to pursue a different approach to install an oxygen functionality at the benzylic positon of our piperidine intermediates. Various reports from the literature have

(82) “Organic Synthesis with Enzymes. 3. TBADH-catalyzed Reduction of Chloro Ketones. Total synthesis of (+)-(S,S)-(cis-6-methyltetrahydropyran-2-yl)acetic Acid: A Civet Constituent,” Keinan, E.; Seth, K. K.; Lamed, R. J. Am. Chem. Soc. 1986, 108, 3474−3480.

Chapter 4, Page 193

described the benzylic oxidation of aromatic-alkyl compounds to benzylic ketones. We hydrogenated the styrenyl olefins in piperidines 4.3 and 4.5 and utilized the resulting products (4.6 and 4.8) as benzylic oxidation substrates.83 As shown in Scheme 4.4 under various benzylic oxidation conditions the desired products 4.7 and 4.9 were not formed the starting piperidine was recovered in quantitative yield in all cases. Among the various oxidants investigated were m-chloroperoxybenzoic acid,84a potassium permanganate supported on montmorillonite K10,77b periodinane reagents,77c pyridinium chlorochromate,77d and 2,3-dichloro-5,6-dicyano-1,4-benzoquinone.77e

variousconditionsN

CO2Et

O

NMeO

MeNCO2Et

O

NMeO

Me

O

Scheme 4.4

4.6 4.7 not formed

variousconditions

4.8

NCO2Et

HO

4.9 not formed

NCO2Et

HO

O

4.3 Discovery of a Unique Directed Hydroalumination

4.3a Observation of an Unexpected Hydroalumination Product. It was at this juncture that we started considering the possibility that the carbamate moiety in 4.3 was responsible for the lack of reactivity of the styrenyl olefin toward Pd and Rh complexes, (and also toward benzylic oxidation conditions in the case of 4.6 and 4.8) either through an electron withdrawing or steric effect. To address this potential problem we chose to reduce the carbamate group to an N-Me. An additional reason for this decision was that

(83) The reaction conditions for the hydrogenation of 4.3 and 4.5 involved 3 equivalents of sodium bicarbonate, 10% by weight to substrate of Pd/C (10% by weight.), as a 1.0 M concentration of substrate in ethanol at 22 oC under 1atm H2. Piperidine 4.6 is isolated in 80% yield and piperidine 4.8 is isolated in 85% yield. (84) (a) Oxygen,

Chem. Soc. 2001, 123, 3183−3185. (d) “Orthocyclophanes. 1. Synthesis and Characterization of [14]- and [15]orthocyclophanes and Bicyclic Biscyclophanes,” Lee, W. Y.; Park, C. H.; Kim, Y. D. J. Org. Chem.

“Oxidation of Benzylic Methylene Compounds to Ketones with m-Chloroperoxybenzoic Acid and ” Ma, D.; Xia, C.; Tian, H. Tetrahedron Lett. 1999, 40, 8915−8917. (b) “Selective Oxidation of

Alkylarenes in Dry Media with Potassium Permanganate Supported on Montmorillonite K10,” Shaabani, A.; Bazgir, A.; Teimouri, F.; Lee, D. G. Tetrahedron Lett. 2002, 43, 5165−5167. (c) “Selective Oxidation at Carbon Adjacent to Aromatic Systems with IBX,” Nicolaou, K. C.; Baran, P. S.; Zhong, Y-L. J. Am.

1992, 57, 4074−4079. (e) 2,3-Dichloro-5,6-dicyano-1,4-benzoquinone (DDQ) in Acetic Acid, a Convenient New Reagent for the Synthesis of Aryl Ketones and Aldehydes via Benzylic Oxidation,” Lee, H.; Harvey, R. G. J. Org. Chem. 1988, 53, 4587−4589.

Chapter 4, Page 194

the alkaloid targets both possess an N-Me group which we would need to eventually unveil in the course of the synthesis. First, the conversion of primary alcohol-containing piperidine 4.5 to N-Me piperidine 4.10 was explored (eq. 4.2). Treatment of piperidine 4.5 with 1.2 equivalents of lithium aluminum hydride delivers the desired N-Me piperidine 4.10 in 46% yield, but to our astonishment also delivers the unexpected piperidine 4.11 in 45 % yield, in which the styrenyl olefin has been saturated. To our knowledge, this is a chemical transformation without precedence.85 Strained olefins in close proximity to amines and propargylic amines have been shown to undergo directed hydroaluminations; however, no allylic amines of this type have been shown to undergo this type of reaction. This reaction piqued my curiosity and was one of my fondest moments in graduate school. I was fortunate throughout my graduate career in starting with set goals that I was able to bring to fruition, namely to develop new catalytic methods that turned out to be highly selective and efficient, but this discovery was different. My personal feelings are that, as a scientist, one of the most exciting and romantic moments a person can have is when nature presents them with a gift such as this. I felt that nature, at that moment in my life, had enticed me to pursue and explore its never-ending secrets.

N

NMe

OH

O

OH

4.11 45%

CO2Et

1.2 equiv. LAH

THF,22 oC, 4 h

H 4.10 46%

4.5NMe

+ (4.2)

4.3b Optimization and Deuterium Labeling Experiments of the Hydroalumination Reaction. The first question to be answered was whether it was possible to perform this reduction reaction under conditions to deliver 4.11 as the major product (Scheme 4.5, see below). Treatment of piperidine 4.5 with two equivalents of lithium aluminum hydride at 65 oC provides N-Me piperidine 4.11 in quantitative yield. The next issue to be (85) For amine directed hydroalumination reactions, see: (a) “Anomalous Lithium Aluminum Hydride Reduction of Carbon—Carbon Double Bonds in 7-azabicyclo(2.2.1)heptenyl Systems,” Marchand, A. P.; Allen, R. W. Tetrahedron Lett. 1975, 16, 67−70. (b) “An Unusual Lithium Aluminum Hydride Reduction of an Isolated Olefin in a 1,2-diamine System,” Szmuszkovicz, J.; Musser, J. H.; Laurian, L. G. Tetrahedron Lett. 1978, 19, 1411−1412. (c) “Stereoselective trans-reduktion tert. Propargylamine mit DIBAH zu (E)-Allylaminen,” Granitzer, W.; Stütz, A. Tetrahedron Lett. 1979, 20, 3145−3148.

Chapter 4, Page 195

addressed was whether this reduction was regioselective. It was still unknown if an organoalane or radical carbon intermediate was being formed in the reaction. To answer this, deuterium labeling experiments were conducted by quenching the reaction with >99.9% deuterated methanol. In doing so, we found that N-Me piperidine 4.12 is formed in 95% yield as a single regioisomer and appears to be a single diastereomer, as determined by 2H NMR studies. The fact that a single regioisomer is formed in the reaction suggests that the reducing agent is directed and that a stabilized C−Al bond forms in the course of the reaction.

NCO2Et

NMe

OH OH

4.11 >98%

2 equiv. LAH

THF,65 C, 2 h H

H

OH D2 equiv. LAH

Scheme 4.5

4.5

NCO2Et

OH

o

NMe

4.12 95%single regioisomer

THF,65 oC, 2 h;MeOD

4.5

4.3c Investigations into Directing Group-Effect of Hydroalumination. The next question asked was what functional group is responsible for the directing effect in this reaction; is the hydroalumination of the styrenyl olefin being directed by the hydroxy group or the amino group? To address the former part of this question, we investigated the reduction of piperidine 4.4, which lacks a primary alcohol (Scheme 4.6, see below). To this end, exposure of 4.4 to lithium aluminum hydride at 65 oC provides N-Me piperidine 4.13 in quantitative yield. This result made it clear that the hydroxy group is importa

wherein the styrenyl olefin h

nt for the reduction of the styrenyl olefin, but it did not negate the requirement for the amino group to participate. For this reason we studied the reduction of pyran 4.14,86 which contains a primary alcohol and styrenyl olefin (Scheme 4.6). Treatment of pyran 4.14 with two equivalents of lithium aluminum hydride at 65 oC provides pyran 4.15,

as been reduced, in quantitative yield. Together, these results provide strong evidence that the hydroxy group and not the amino group is

(86) This compound was prepared from an intermediate that is discussed in Chapter 5.

Chapter 4, Page 196

responsible for this unique directed hydroalumination reaction.87 It is interesting to note that this reaction can be applied to pyrans as well as piperidines and, although not studied, it would be curious to know whether this reaction could be applied to other six-membered ring hetero- or carbocycles.

NCO2Et

2 equiv. LAH

Scheme 4.6

NMe

4.13 >98%

THF,65 oC, 2 h

4.4

O O

4.15 >98%

2 equiv. LAH

THF,65 oC, 2 h

4.14

MeMeOBn

MeOBn

Me

HO HO

4.3d Investigations into Trapping of Organoalane Intermediate. After these mechanistic investigations were completed, we focused our, efforts on performing a directed hydroalumination reaction of piperidine 4.5 followed by trapping of the organoalane intermediate formed in the reaction with an electrophilic oxygen source. The natural products lobeline and sedinone both possess a benzylic alcohol functionality and we felt this would be a unique way of introducing this functional group to this class of piperidines. And we came to the conclusion that the best electrophilic oxygen source would be anhydrous oxygen gas, given practicality and atom economy of using this reagent. The outcome of these studies is shown in eq. 4.3. To this end, treatment of piperidine 4.5 with two equivalents of lithium aluminum hydride at 65 oC followed by exposure of the resulting mixture to anhydrous oxygen gas provides N-Me piperidine 4.16 in 55% yield and N-Me 4.11 in 40% yield. We were content that the desired product 4.16, wherein oxygen is incorporated at the benzylic position, is formed in the reaction. (87) For a review of substrate-directed reactions, see: (a) “Substrate-directable Chemical Reactions,” Hoveyda, A. H.; Evans, D. A.; Fu, G. C. Chem. Rev. 1993, 93, 1307−1370. For examples of hydroxyl-directed hydroalumination reactions, see: (b) “Stereochemical Control of Reductions. 6. The Hydroxymethyl Group as a Hinge for Internal Reagent Delivery,” Thompson, H. W.; McPherson, E. J. Org. Chem. 1977, 42, 3350−3353. (c) “α-Ketoketene Dithioacetal Chemistry. 1. Alternate Modes of Lithium Aluminum Hydride Reduction. Regio- and Stereospecific vs. Reduction-alkylation-fragmentation,” Gammill, R. B.; Gold, P. M.; Mizsak, S. A. J. Am. Chem. Soc. 1980, 102, 3095−3100. (d) “Hydroxy-directed Hydroaluminations: A Stereoselective Approach to Cycloalkanols From β-aryl Enones,” Tetrahedron Lett. 1994, 35, 1137−1140.

Chapter 4, Page 197

Unfortunately, the reaction is not stereoselective and 4.16 is isolated as a 1:1 mixture of epimers at the benzylic position. This result was in contrast to our previous deuterium labeling studies, in which the product appears to be a single diastereomer (see eq. 4.2 above). It is probable that in the formation of 4.16 oxygen is incorporated via a radical mechanism. We also became concerned with incomplete oxygen incorporation in the reaction and under a variety of conditions were not able to prevent the formation of 4.11.

NCO2Et

OH2 equiv. LAH

THF, 65 oC, 2 h;O2

NMe

4.16 55% 1:1 dr

OH OH

NMe

4.11 40%

OH

+

4.5

(4.3)

To nullify the possibility that an adventitious proton source was responsible for the formation of 4.11, we performed the electrophilic oxygen trapping reaction with a deuterium quench (eq. 4.4). Treatment of piperidine 4.5 with lithium aluminum hydride at 65 oC, exposure of the resulting mixture to anhydrous oxygen gas and then treatment with >99.9% deuterated methanol again deliverer N-Me piperidine 4.16 in 55% yield and the deuterium-incorporated product 4.12 in 40% yield. This result provided proof that an advantitious proton source is not responsible for formation of 4.11. Most likely, organoaluminum intermediates aggregate during the course of the reaction, which prevents full oxygen incorporation at the benzylic position.

NCO2Et

OH2 equiv. LAH

THF, 65 oC, 2 h;

N

4.16 55%

OH OH

OH

+

D(4.4)

O2, MeOD

Me

NMe

4.12 40%

4.5

Chapter 4, Page 198

4.4 Alternative ApproachTo Piperidine Alkaloids

4.4a Tandem Alkylation-Reduction/Hydroalumination-Oxygen Trapping of Piperidine Intermediates. Despite being unable to fully incorporate oxygen in the directed hydrometallation of 4.5, we were optimistic that we could still use this unique reaction and devised a new strategy toward the alkaloids lobeline and sedinone (Scheme 4.7).

NCO2Et

O

NMeO

Me

i) MeMgClor PhMgCl

ii) LAH

ClMg O

ClMg

NMe

R

O OH

R = Ph, lobelineR = Me, sedinone

4.3

one pot

NCO2Et

O

NMeO

Me

4.3

Scheme 4.7

NCO2Et

OH

4.5

NCO2Et

N

OMeO

MeR

NMe

NMeO

MeR

(RO)2Al

iii) O2

4.17 Formation of StableTetrahedral Intermediate

4.18 Directed OlefinReduction

Trapping

This approach involved a multi-component, one-pot cascade of reactions starting from piperidine 4.3. Due to the stability of Weinreb amide tetrahedral intermediates, we hypothesized that alkylation of the Weinreb amide in 4.3 could be performed with either methyl or phenyl nucleophile, resulting in the likely robust intermediate 4.17. Then, eatment of this intermediate with lithium aluminum hydride would result in the

eduction of the carbamate in 4.17, followed by the directed hydroalumination to generate

atrr

Chapter 4, Page 199

organoalane intermediate 4.18. Treatment of this intermediate with anhydrous oxygen

ining the feasibility of performing an alkylation of the Weinreb amide in 4.3 followed by reduction of the ethyl-carbamate group to an N-Me group (Scheme 4.8). Toward this objective, treatment of piperidine 4.3 with 1.1 equiv. of methylmagnesium chloride followed by treatment with 1 equiv. of lithium aluminum hydride at 65 oC delivers the desired piperidine 4.19 in 75% yield after 1 h. If more than 1 equiv. of lithium aluminum hydride is used in the reaction, undetermined side products begin to form. Additionally, the stereogenic center β to the methyl ketone undergoes a facile epimerization resulting in a mixture of 2,6-syn and 2,6-anti piperidines.88 To ensure the fidelity of the directed hydrometallation we performed the reaction with a deuterium quench and were able to isolate the desired product 4.20 in 64% yield. Next, we pursued the multi-component reaction sequence with the inclusion of an oxygen gas trapping after the alkylation/directed hydrometallation sequence. Unfortunately, under a variety of reaction conditions, desired natural products were not formed; analysis of the reaction mixture by 1H NMR analysis revealed that complex mixtures formed in all cases. These studies suggest that introduction of oxygen results in break down of the tetrahedral intermediate formed from the Weinreb amide.

followed by standard work-up conditions was then expected to furnish either lobeline or sedinone.

4.4b Studies into Multicomponent Reactions of a Weinreb Amide-containing Piperidine. Investigations into this multi-component transformation began with determ

NCO2Et

O

NMeO

Me

1.1 equiv. MeMgCl;1 equiv. LAH

THF, 65 oC, 1h;MeOH (MeOD)

NMe

4.19 75% (H)4.20 64% (D)

O H(D)

N

O

NMeO complex mixture

1.1 equiv. MeMgCl;1 equiv. LAH

o

4.3

Scheme 4.8

CO2EtMe THF, 65 C, 1h;O2

4.3

(88) It is well documented that β-amino ketones such as 4.19 undergo facile epimerization, presumably through an elimination/intramolecular conjugate addition process, see: “Syntheses of the Sedum and Related Alkaloids,” Bates, R. W.; Sa-Ei, K. Tetrahedron 2002, 58, 5957−5978.

Chapter 4, Page 200

4.5 Miscellaneous Functionalization Reactions of Piperidines

Prior to concluding this chapter several other piperidine funtionalizations will be discussed. Most importantly, the functionalizations performed to determine the absolute stereochemistry of the piperidine products in the Mo-catalyzed asymmetric ring-opening/cross-metathesis reactions of azabicycles will be discussed.

4.5a Absolute Stereochemistry Determination for Piperidine Products. The absolute stereochemistry for piperidine products was determined by chemical correlation to a piperidine product of known absolute stereochemistry (Scheme 4.9).89 The stereoproof began with piperidine 4.5 of 80% ee.90 Oxidation of 4.5 to the corresponding aldehyde with PCC, followed by alkylation with diphenylzinc provides a mixture of alcohols 4.21 and 4.22 that is separable by silica gel chromatography. Treatment of piperidine 4.22 with lithium aluminum hydride then affords piperidine 4.23,91 of known absolute stereochemistry. The calculated enantiomeric excess for a sample of 4.23 of [αD]20 = +15.6 (c = 0.1, CHCl3) is 76% ee, correlating to our observed value of 80% ee. The absolute stereochemistry of all piperidine products was assigned by inference based on the absolute stereochemistry of piperidine 4.5, which was obtained from the enantioenriched piperidine 4.4.

4 Å MS, CH2Cl2, 2 h

Scheme 4.9

OH

NCO2Et

Ph

OH

Ph

1) 1.5 equiv. PCC, NaOAc

2) 3 equiv. Ph2Zn,Et2O

22 oC, 18 h

NCO2Et

Ph

4.5 (80% ee)

4.21

NCO2Et

Ph

OH

Ph

4.22

+

NCO2Et

Ph

OH

Ph

4.22

2 equiv. LAHTHF, 65 oC, 2 h

NMe

Ph

OH

Ph

4.23 [α]D20 = +15.6 (c = 0.1, CHCl3)

ref. 82 [α]D25 = +29.6 (c = 1.0, CHC ) l3

(89)“Lobeline: Structure-Affinity Investigation of Nicotinic Acetylcholinergic Receptor Binding,” Flammia, D.; Dukat, M.; Damaj, M. I.; Martin, B.; Glennon, R. A. J. Med. Chem. 1999, 42, 3726−3731. (90) In the optimized reaction conditions for AROM/CM, piperidine 4.4 was obtained in 90% ee, see Ch 3. (91) The spectral data for 4.23 has been previously published, see ref. 82.

Chapter 4, Page 201

4.5b Epoxidation of Substituted Styrene Intemediate. One functionalization of

could be performed

these piperidine products involved the epoxidation of the styrenyl olefin in piperidine 4.3 (eq. 4.5). Treatment of piperidine 4.3 with 1.5 equivalents of m-chloroperoxybenzoic acid in methylene chloride delivers expoxide 4.24 in 64% yield as a 1:1 mixture of epoxide diastereomers. It was hoped that the opening of the epoxideregioselectively with a hydride source through a directing effect from the ethyl-carbamate en route towards the total synthesis of lobeline and sedinone. Unfortunately, several hydride sources resulted in hydride attack at the benzylic position due to electronic stabilization.

NCO2Et

O

NMeO

Me

1.5 equiv. m-CPBA

4.3

NCO2Et

O

NMeO

Me

4.24 64%

O(4.5)

CH2Cl2,0 oC to 22 oC

2 h

4.5c Dissolving Metal Vinylogous Opening of an N-Me Piperidine to a 1,3-Aminoalcohol. Another functionalization of piperidine products involved the conversion of N-Me piperidine 4.25 to 1,3-aminoalcohol 4.26 (eq. 4.6). Previously, we described the dissolving metal opening of related pyrans to 1,3-diols.92 We found this reaction can also be applied to styrenyl olefin containing piperidines. Treatment of piperidine 4.25 with sodium dissolved in liquid ammonia delivers 1,3-aminoalcohol 4.26 in 85% yield after 15 min.

NMe

Ph

OTBS

Na(0), NH3

MeHN OTBS

4.26 85%

t-BuOH, − 78 oC,15 min

4.25

(4.6)

4.6 Conclusions

This chapter described several methods for the functionalization of enantioenriched piperidine products that were derived from Mo-catalyzed asymmetric ring-opening/cross-metathesis reactions of azabicycles. Initially, efforts were directed toward the total synthesis of piperidine alkaloids and it was shown that terminal olefins can be easily differentiated from styrenyl olefins. The difficulties encountered in functionalizing the styrenyl olefin of piperidine intermediates underlines the need for new

(92) “Efficient Enantioselective Synthesis of Functionalized Tetrahydropyrans by Ru-Catalyzed

. Asymmetric Ring-Opening Metathesis/Cross-Metathesis (AROM/CM),” Gillingham, D. G.; Kataoka, O.; Garber, S. B.; Hoveyda, A. H. J. Am. Chem. Soc. 2004, 126, 12288−12290

Chapter 4, Page 202

method

AROM/CM reactions.

4.7 Experimental Section

Synthesis of primary alcohol-containing piperidine 4.5. A 25-mL round-bottom flask was charged with piperidine 4.493 (0.250 g, 0.876 mmol), THF (9.0 mL), 9-borabicyclo[3.3.1]nonane (0.130 g, 1.10 mmol, 1.25 equiv.), and allowed to stir for 1 h. At this time, the reaction mixture

was cooled to −10 oC in a NaCl/ice bath, and the reaction was quenched by the (slowly and sequentially) addition of a 30% aqueous solution of H2O2 (0.880 mL, 1 mL/mmol of 4.4), and a 1.0 M solution of NaOH (0.880 mL, 1 mL/mmol of 4.4). The resulting mixture was warmed to 22 oC over 30 min, after which time it was transferred to a 60-mL separatory funnel. At this point, EtOAc (15 mL), and a saturated aqueous solution of NaCl (10 mL) were transferred to the separatory funnel, and the resulting biphasic layers were separated. The organic layer was washed with H2O (2 × 10 mL), dried (MgSO4), filtered, and the volatiles were removed in vacuo. The resulting colorless residue was purified by silica gel chromatography (1:1 EtOAc:hexanes) to deliver 4.5 as colorless oil (0.258 mg, 0.850 mmol, 97%). IR (neat): 3446 (br), 2943 (m), 2867 (m), 1659 (s), 1413 (m). 1H NMR (400 MHz, CDCl3): δ 7.30−7.20 (m, 5H), 6.52 (d, J = 16.0 Hz, 1H), 6.19 (dd, J = 16.0, 6.8 Hz, 1H), 5.00−4.92 (br, 1H), 4.52−4.44 (m, 1H), 4.28−4.14 (m, 2H), 3.62−3.42 (m, 2H), 1.98−1.55 (m, 8H), 1.25 (t, J = 9.6 Hz, 3H). 13C NMR (100 MHz, CDCl3): δ 171.1, 136.8, 130.8, 130.1, 128.6, 127.6, 126.2, 61.9, 60.4, 58.8, 51.1, 46.7,

m/z) Calcd for C18H25NO3 303.1834

Synthesis of piperidine 4.3. A 25-mL round-bottom flask was charged with 4.5 (317 mg, 1.05 mmol), CH2Cl2 (10.5 mL), NaOAc (85.0 mg, 1.05 mmol, 1

s to be developed for reactions of substituted styrenes of molecular complexity. Nevertheless, these investigations resulted in the discovery and development of a novel regioselective hydroalumination reaction of styrenyl olefins. Various studies, including deuterium labeling experiments, were conducted to shed insight into the mechanism of this reaction. Furthermore, it was shown that this novel transformation can be employed as a benzylic oxidant by the trapping of an organoalane with an electrophilic oxygen source. Another set of important funcionalization reactions resulted in the preparation of a piperidine that established the absolute stereochemistry of piperidines derived from Mo-catalyzed

NCO Et

OH

2

4.5

37.1, 29.4, 28.3, 21.0, 14.8, 14.7, 14.2. HRMS EI ((M)+, Found 303.1833.

OMeO

(93) For the characterization data of piperidine 4.4 please see the experimenta

2

4.3MeN N

CO Et

l section of Ch 3.

Chapter 4, Page 203

equiv.), powdered 4Å molecular sieves (15.9 mg, 5 wt. %), PCC (339 mg 1.58 mmol, 1.5 equiv.), and the mixture was allowed to stir for 2 h. To the reaction mixture was added Et2O (5.0 mL), and resulting mixture was allowed to stir for 15 min. At this time, the mixture was filtered through Florisil® by vacuum filtration, eluted with EtOAc, and the filtrate was collected in a 25-mL round-bottom flask. The volatiles were removed in vacuo. The resulting slightly yellow residue was dissolved in t-BuOH (5.3 mL), and tetramethylethylene (936 µL, 7.88 mmol, 7.5 equiv.). To this solution was added a solution of NaClO2 (712 mg, 7.88 mmol, 7.5 equiv.), and NaH2PO4 (942 mg, 6.83 mmol, 6.5 equiv.) dissolved in H2O (5.3 mL). The reaction mixture was allowed to stir for 1 h,

s were removed in vacuo. To this mixture was added EtOAc (10 mL), H2O (10 mL), thresulting biphasic layers waqueous solution of NaCconcentrated to yellow oilmL round-bottom flask,

RMS EI (m/z) mass Calcd for C20H28N2O4Na 383.1947 (M+Na) , Found 383.1928.

Synthesvial waTHF (0.500 mL), and lithium aluminum hydride (6.0 mg,

after which time the volatilee mixture was transferred to a 60-mL separatory funnel, and the ere separated. The organic layer was washed with a saturated l (10 mL), H2O (10 mL), dried (Na2SO4), filtered, and

. This oil was dissolved in minimal EtOAc, transferred to a 25-and re-concentrated under reduced pressure. The resulting

residue was dissolved in CH2Cl2 (10.5 mL). To this solution was added Et3N (295 µL, 2.10 mmol, 2 equiv.), EDC (301 mg, 1.58 mmol, 1.5 equiv.), HOBt (241 mg, 1.58 mmol, 1.5 equiv.), N,O-dimethylhydroxylamine hydrochloride (154 mg, 1.58 mmol, 1.5 equiv.), and the mixture was allowed to stir for 2 h. At this time, the reaction was quenched by the addition of a saturated aqueous solution of NH4Cl (10 mL), the mixture was transferred to a 60-mL separatory funnel, and the resulting biphasic layers were separated. The organic layer was washed with H2O (2 x 10 mL), dried (Na2SO4), and the volatiles were removed in vacuo. The resulting yellow residue was purified through silica gel chromatography (1:1 EtOAc:hexanes) to furnish 4.3 as colorless oil (295 mg, 0.819 mmol, 78% over three steps). IR (neat): 2936 (s), 1690 (s), 1407 (m), 1312 (m), 1268 (m). 1H NMR (400 MHz, CDCl3): δ 7.39−7.22 (m, 5H), 6.53 (d, J = 16.4 Hz, 1H), 6.32 (dd, J = 16.4, 5.2 Hz, 1H), 5.01 (br, 1H), 4.78 (br, 1H), 4.24−4.10 (m, 2H), 3.52 (s, 3H), 3.13 (s, 3H), 2.96−2.84 (m, 1H), 2.55 (dd, J = 14.8, 3.6 Hz, 1H), 2.05−2.0 (m, 1H), 1.82−1.60 (m, 5H), 1.30−1.22 (m, 3H). 13C NMR (100 MHz, CDCl3): δ 156.0, 137.5, 137.2, 131.4, 130.1, 128.7, 127.6, 126.3, 61.5, 61.3, 60.5, 50.8, 47.5, 36.1, 32.2, 27.8, 21.2, 14.9, 14.6, 14.3. H +

is of piperidine 4.12. In a N2-filled glovebox, a 4-mL s charged with piperidine 4.5 (15.0 mg, 0.0494 mmol), N

Me

OH

4.12

D

Chapter 4, Page 204

0.099 mmol, 2 equiv.). The vial was tightly sealed with a Teflon cap, placed in a heating mantle at 65 oC, and the reaction mixture was allowed to stir for 1 h. At this point, the reaction mixture was cooled to 22 oC over 15 min, the vial was removed from the glovebox, and the reaction was quenched by the slow addition of MeOD94 (0.500 mL). The resulting mixture was allowed to stir for 12 h. To this mixture was (slowly and sequentially) added H2O (15.0 µL, 1 µL/mg of 4.5), a 3.8 M solution of NaOH (45.0 µL, 3 µL/mg of 4.5), and H2O (15.0 µL, 1 µL/mg of 4.5). The resulting mixture was subjected to vacuum filtration (to remove Al salts), and eluted with EtOAc. The filtrate was dried (Na2SO4), re-filtered, and the volatiles were removed in vacuo. The resulting colorless residue was purified by neutral alumina gel chromatography (20:1 CH2Cl2:MeOH) to provide piperidine 4.12 as colorless oil (11.8 mg, 0.0475 mmol, 95%). IR (neat): 3371 (br), 2930 (s), 2855 (m), 1055 (w). 1H NMR (400 MHz, C6D6): δ 7.20−7.10 (m, 5H), 3.92 (ddd, J = 10.4, 5.2 Hz, 1H), 3.86−3.79 (m, 1H), 2.52−2.42 (m, 2H), 2.30−2.20 (m, 1H), 2.02 (s, 3H), 1.59−0.86 (m, 8H). 1H NMR (400 MHz, CDCl3, TMS): δ 7.30−7.18 (m, 5H), 3.92 (ddd, J = 10.8, 5.2 Hz, 1H), 3.86−3.79 (m, 1H), 2.82−2.76 (m, 1H), 2.62−2.50 (m, 2H), 2.27 (s, 3H), 1.82−1.24 (m, 8H). 13C NMR (100 MHz, C6D6): δ 143.0, 129.0, 128.9, 126.3, 64.4, 62.9, 62.5, 36.3, 33.9, 32.1, 31.9, 31.7, 30.0, 26.2, 26.0, 25.4. HRMS EI (m/z) Calcd for C16H25DNO 249.2077 (M+H)+, Found 249.2015. 2H NMR (400 MHz, CDCl3): δ 2.50−2.40 (br s).

Synthesis of piperidine 4.11. The procedure to synthesize piperidine 4.11 was identical to that used for the synthesis of 4.12, however, MeOH was used to quench the reaction mixture. IR (neat): 3390 (br), 2930 (s), 2855 (m), 1457 (w),

1055 (w). 1H NMR (400 MHz, C6D6): δ 7.30−7.18 (m, 5H), 3.92 (ddd, J = 10.8, 5.2 Hz, 1H), 3.86−3.79 (m, 1H), 2.62−2.50 (m, 3H), 2.31−2.25 (m, 1H), 2.00 (s, 3H), 1.70−1.60 (m, 1H), 1.59−1.21 (m, 5H), 1.17−0.80 (m, 4H). 1H NMR (400 MHz, CDCl3): δ 7.30−7.25 (m, 3H), 7.20−7.14 (m, 2H), 3.92 (ddd, J = 10.4, 5.2, 5.2 Hz, 1H), 3.86−3.79 (m, 1H), 2.84−2.76 (br, 1H), 2.68−2.55 (m, 3H), 2.02 (s, 3H), 1.85−1.72 (m, 3H), 1.70−1.60 (m, 3H), 1.58−1.40 (m, 3H), 1.30−1.20 (m, 2H). 13C NMR (100 MHz, CDCl6): δ 142.6, 128.5, 125.8, 64.6, 63.2, 62.6, 36.1, 32.6, 32.0, 26.0, 25.7, 24.9. HRMS EI (m/z) Calcd for C16H26NO 248.2014 (M+H)+, Found 248.2015.

NMe

OH

4.11H

H

Synthesis of piperidine 4.13. The procedure to synthesize p4

iperidine 4.13 was identical to that used for the synthesis of .12, however, MeOH was used to quench the reaction mixture.

(94) Deuterated methanol was t

NMe

4.13aken from a freshly opened ampule of 99.95 atom % D.

Chapter 4, Page 205

IR (neat): 2930 (s), 2848 (m) 2773 (s), 1445 (w). 1H NMR (400 MHz, CDCl3): δ 7.40−7.20 (m, 5H), 6.49 (d, J = 16.0 Hz, 1H), 6.18 (dd, J = 8.8, 7.2 Hz, 1H), 5.82−5.73 (m, 1H), 5.15 (dd, J = 17.6, 2.0 Hz, 1H), 5.03 (dd, J = 10.4, 1.6 Hz, 1H), 2.60−2.54 (m, 1H), 2.47−2.42 (m, 1H), 2.21 (s, 3H), 1.80−1.42 (m, 6H). 13C NMR (100 MHz, CDCl3): δ 142.7, 137.4, 134.3, 130.5, 128.8, 127.5, 126.4, 115.5, 69.4, 68.5, 42.3, 33.9, 33.8, 24.0. HRMS EI (m/z) Calcd for C16H21N 227.1674 (M)+, Found 227.1679.

Synthesis of piperidine 4.15. The procedure to synthesize piperidine 4.15 was identical to that used for the synthesis of 4.12, however, MeOH was used to quench the reaction mixture.

Synthesis of piperidine 4.16. In a N -filled glovebox, a 4-mL vial was charged with piperidine 4.5 (40.0 mg, 0.130 mmol), THF (1.30 mL), and lithium aluminum hydride (10.0 mg, 0.260 mmol, 2 equiv.). The vial was tightly sealed with a Teflon cap,

placed in a heating mantle at 65 oC, and the mixture was allowed to stir for 1 h. At this moment, the reaction mixture was cooled to 22 oC over 15 min, the vial was removed from the glovebox, and O (balloon passed through drying tube filled with P O ) was bubbled through the reaction mixture for 1 h. To this mixture were slowly and sequentially added H O (40.0 µL, 1 µL/mg of 4.5), a 3.8 M solution of NaOH (0.120 mL, 3 µL/mg of 4.5), and H O (40.0 µ

2

2 2 5

2

2 The resulting mixture was subjected to vacuum filtrawas dried (Na2SO4), re-filcolorless residue was CH2Cl2:MeOH) to provide

µL, 0.0610 mmol, 1.1 equiv.) was added to this solution by

L, 1 µL/mg of 4.5).

NMe

4.16

OH OH

tion (to remove Al salts), and eluted with EtOAc. The filtrate tered, and the volatiles were removed in vacuo. The resulting purified by neutral alumina gel chromatography (10:1 piperidine 4.16 as colorless oil (18.8 mg, 0.0715 mmol, 55%).

IR (neat): 3370 (br), 2930 (s), 2855 (m), 1451 (w), 1055 (w). Isolated as a 1:1 mixture of diastereomers. 1H NMR (400 MHz, CDCl3, TMS): δ 7.39−7.23 (m, 5H), 5.00 (dd, J = 6.0, 6.0 Hz, 0.5H), 4.92 (dd, J = 10.0, 3.6 Hz, 0.5H), 3.82−3.64 (m, 2H), 3.14−3.00 (m, 1H), 2.84−2.60 (m, 1H), 2.34 (s, 1.5H), 2.28 (s, 1.5H), 2.00−1.20 (m, 10H). 13C NMR (100 MHz, CDCl3): δ 145.9, 145.4, 128.8, 127.6, 127.4, 126.2, 126.0, 75.6, 72.2, 64.1, 63.2, 62.9, 62.6, 61.6, 61.3, 61.0, 60.9, 41.0, 40.8, 35.8, 35.4, 32.8, 30.3, 30.2, 26.7, 25.4, 25.3, 25.0, 24.2, 24.0. HRMS EI (m/z) mass Calcd for C16H25NO2 263.1885 (M)+, Found 263.1883.

Synthesis of piperidine 4.20. A 4-mL vial was charged with 4.3 (20.0 mg, 0.0555 mmol), and THF (555 µL). Methylmagnesium chloride (a 2.40 M solution in THF, 25.4

NMe

4.20

O D

Chapter 4, Page 206

syringe, and the mixture was allowed to stir for 30 min. At this time, the vial was relocated into a N2-filled glovebox. To this mixture was added lithium aluminum hydride (2.1 mg, 0.055 mmol, 1 equiv.), the vial was tightly sealed with a Teflon cap, placed in a heating mantle at 65 oC, and allowed to stir for 1 h. At this point, the mixture was cooled to 22 oC over 15 min, the vial was removed from the glovebox, and the reaction was quenched by the slow addition of MeOD5 (500 µL). The resulting mixture was allowed

4.3). The mixture was filtered under vacuum (to remove Al salts), and eluted with EtOAc. The resultinremoved in vacuo. The chromatography (20:1 CHmg, 0.0415 mmol, 75%).9

was quenched by the addition of a saturated aqueous solution ow

to stir for 12 h. To this mixture were slowly and sequentially added H2O (20 µL, 1 µL/mg of 4.3), a 3.8 M solution of NaOH (60 µL, 3 µL/mg of 4.3), and H2O (20 µL, 1 µL/mg of

g filtrate was dried (Na2SO4), re-filtered, and the volatiles were resulting yellow residue was purified by neutral alumina gel 2Cl2:MeOH) to provide piperidine 4.20 as colorless oil (10.8

5 IR (neat): 2930 (s), 2855 (s), 1709 (s), 1451 (m). 1H NMR (400 MHz, CDCl3, TMS): δ 7.30−7.17 (m, 5H), 2.70−2.40 (m, 4H), 2.32 (s, 3H), 2.15 (s, 3H), 1.80−1.25 (m, 9H). 13C NMR (100 MHz, CDCl3): δ 142.2, 133.6, 128.6, 128.5, 128.4, 125.9, 62.8, 59.6, 56.2, 55.0, 52.5, 50.5, 38.9, 38.1, 36.0, 32.3, 32.2, 31.8, 29.8, 27.4, 26.9, 24.7, 24.6, 19.7. HRMS EI (m/z) mass Calcd for C17H24NOD 260.2018 (M)+, Found 260.2014. 2H NMR (400 MHz, CDCl3): δ 2.16 (s), 2.14 (s).

Piperidine 4.19.92 IR (neat): 2923 (s), 2855 (s), 1709 (s), 1602 (m), 1451 (m). 1H NMR (400 MHz, CDCl3): δ 7.30−7.22 (m, 2H), 7.20−7.15 (m, 3H) 2.70−2.40 (m, 4H), 2.34 (s, 3H), 2.20 (s, 3H), 2.16 (d, J = 7.2 Hz, 2H), 1.80−1.25

(m, 8H). 13C NMR (100 MHz, CDCl3): δ 140.6, 129.1, 128.9, 128.8, 126.7, 57.4, 54.1, 47.5, 36.8, 32.1, 30.8, 30.1, 29.8, 23.8, 23.6. HRMS EI (m/z) mass Calcd for C17H25NO 259.1936 (M)+, Found 259.1935.

Synthesis of piperidines 4.21 and 4.22. In a N2-filled glovebox, diphenylzinc (22.0 mg, 0.0990 mmol, 3 equiv.) was added to a solution of the aldehyde derived from 4.5 (10.0 mg, 0.0330 mmol, 1 equiv.) in Et2O (0.400 mL) and the reaction mixture was allowed to stir for 18 h. At this time, the reaction

NMe

4.19

O H

NCO2Et

Ph

OH

Ph

4.21

OH

Ph NCO2Et4.22

Phf NaHCO3 (2 mL) and diluted with Et2O (2 mL), the mixture as transferred to a 30-mL separatory funnel, and the

ixture of 2,6-syn and anti isomers, resulting in 13C NMR spectra that does

(95) This compound exists as a mnot appear clean, see ref. 85 of this Chapter.

Chapter 4, Page 207

resulting biphasic layers were separated. The organic layer was washed with H2O (2 x 10 mL), dried (Na2SO4), and the volatiles were removed in vacuo. The resulting colorless residue was purified through silica gel chromatography (1:5 EtOAc:hexanes) to furnish 4.21 as colorless oil (5.0 mg, 0.013 mmol, 40%, TLC Rf = 0.35) and 4.22 as colorless oil (5.0 mg, 0.013 mmol, 40%, TLC Rf = 0.2). 1H NMR of 4.21 (400 MHz, CDCl3): δ 7.39−7.27 (m, 5H), 7.13−7.09 (m, 5H), 6.65 (d, J = 15.6 Hz, 1H), 6.38 (dd, J = 16.0, 6.8 Hz, 1H), 5.08−4.26 (m, 3H), 4.24−4.18 (m, 2H), 2.20 (dd, J = 12.8, 1.6 Hz, 1H), 2.00−1.90 (m, 1H), 1.84−1.70 (m, 2H), 1.66−1.1.50 (m, 2H), 1.36 (t, J = 7.2 Hz, 3H), 0.90−0.78 (m, 2H). 1H NMR of 4.22 (400 MHz, CDCl3): δ 7.36−7.21 (m, 10H), 6.51 (d, J = 16.0 Hz, 1H), 6.24 (dd, J = 16.0, 7.2 Hz, 1H), 4.96−4.86 (m, 1H), 4.70−4.64 (m, 1H), 4.46−4.40 (m, 1H), 4.25−4.13 (m, 2H), 2.26−1.90 (m, 2H), 1.80−1.70 (m, 2H), 1.60−1.54 (m, 2H), 1.34−1.22 (m, 3H), 0.90−0.80 (m, 2H).

Synthesis of epoxide 4.24. A 5-mL round-bottom flask was charged with piperidine 4.3 (27.0 mg, 0.0750 mmol, 1 equiv.), CH2Cl2 (0.750 mL), and the reaction vessel was cooled to 0 oC in an ice-bath. To this

mixture was slowly added m-chloroperbenzoic acid (20.0 mg, 0.110 mmol, 1.5 equiv.). At this point the flask was removed from the ice-bath, the reaction

NCO2Et

O

NMeO

Me4.24

O

mixture was allowed to stir at 22 oC for 2 h afterAt this time H2O (5 mL) andthe resulting biphasic layers(3 × 10 mL). The comb

oxide diastereomers). 1H NMR (400 MHz, CDCl3):96 δ 7.33−7.21 (m,0.5H), 4.20−4.08 (m, 2H),

a dry ice/acetone bath. Adissolution of Na in liquid ammonia the solution became deep blue in appearance. A

which time it was transferred to a 30-mL separatory funnel. CH2Cl2 (5 mL), was transferred to the separatory funnel and

were separated. The aqueous layer was washed with CH2Cl2 ined organic layers were dried (MgSO4), filtered, and the

volatiles removed in vacuo. The resulting colorless residue was purified through silica gel chromatography (2:3 EtOAc:hexanes) to deliver 4.24 as colorless oil (18.0 mg, 0.0478 mmol, 64%, mixture of circa 1:1 exp

5H), 4.80−4.70 (m, 1H), 4.57−4.52 (m, 0.5H), 4.32−4.26 (m, 3.93 (s, 0.5H), 3.78 (s, 0.5H), 3.44 (s, 3H), 3.18−3.02 (m, 4H), .42 (m, 1H), 1.90−1.50 (m, 6H), 1.30−1.20 (m, 3H).

Synthesis of 1,3-aminoalcohol 4.26. A 25-mL round-bottom flask was charged with Na (40.0 mg, 1.74 mmol, 48 equiv.), the flask was fitted with a cold-finger condenser, and the reaction vessel was cooled to –78 oC in

mmonia gas (circa 5 mL) was condensed into the flask; upon

2.90−2.78 (m, 1H), 2.58−2

MeHN OTBS4.26

(96) The 1H NMR spectra shown contains residual EtOAc.

Chapter 4, Page 208

solution of piperidine 4.25 (13.0 mg, 0.0360 mmol, 1 equiv.) in t-BuOH (0.120 mL), and Et2O (2.0 mL) was added to this solution by syringe. The reaction mixture was allowed to stir at –78 oC for 15 min, after which time the reaction was quenched by the addition of CH2Cl2 (10 mL), and the cold bath was removed. The resulting mixture was allowed to warm to 22 oC over 15 min, and then transferred to a 60-mL separatory funnel. At this instance, H2O (10 mL) was transferred to the separatory funnel, and the resulting biphasic layers were separated. The aqueous layer was washed with CH2Cl2 (3 × 10 mL). The combined organic layers were dried (MgSO4), filtered, and concentrated in vacuo to deliver analytically pure amino alcohol 4.26 as yellow oil (11.0 mg, 0.0304 mmol, 85%). IR (neat): 2949 (s), 2924 (s), 2855 (m), 2773 (w), 1105 (m). 1H NMR (400 MHz, CDCl3): δ 7.30−7.25 (m, 3H), 7.20−7.15 (m, 2H) 5.77 (ddd, J = 17.6, 10.4, 8.4 Hz, 1H), 5.12 (dd, J = 17.6, 0.8 Hz, 1H), 5.04 (dd, J = 10.4, 0.8 Hz), 3.65−3.59 (m, 1H), 2.78−2.70 (m, 1H), 2.61−2.50 (m, 1H), 2.1813C NMR (100 MHz, CDCl3): δ 140.2, 38.5, 35.7, 31.3, 29.8, 26.0C22H37NOSi 359.2644 (M−H2)+, F

(s, 3H), 2.05−1.42 (m, 9H), 0.89 (s, 9H), 0.07 (s, 6H).

28.5, 128.4, 125.9, 115.4, 69.3, 67.3, 66.0, 61.5, 42.0, , 18.3, 15.4, −4.40. HRMS ES (m/z) mass Calcd for ound 359.2647.

Chapter 4, Page 209

Chapter 4, Page 210

Chapter 4, Page 211

Chapter 4, Page 212

Chapter 4, Page 213

Chapter 4, Page 214

Chapter 4, Page 215

Chapter 4, Page 216

Chapter 4, Page 217

Chapter 4, Page 218

Chapter 4, Page 219

Chapter 4, Page 220

Chapter 4, Page 221

Chapter 4, Page 222

Chapter 4, Page 223

Chapter 4, Page 224

Chapter 4, Page 225

Chapter 4, Page 226

Chapter 4, Page 227

Chapter 4, Page 228

Chapter 4, Page 229

Chapter 4, Page 230

Chapter 5, Page 231

Chapter Five

COMPARISON OF MOLYBDENUM AND RUTHENUIM COMPLEXES FOR CATALYTIC ASYMMETRIC RING-OPENING CROSS-METATHESIS REACTIONS

5.1 Introduction

In Chapter Two I alluded to the reactivity difference between Mo and Ru complexes for olefin metathesis. This difference affords chiral Mo complexes an advantage over chiral Ru complexes for the asymmetric ring-closing metathesis of tertiary amines; chiral Ru complexes were inefficient at performing asymmetric ring-closing metathesis reactions of amines. Nonetheless, we did find one example where Mo complexes were inefficient; namely the desymmetrization reaction of N-Bn azabicycles. Would chiral Ru complexes be efficient for this type of transformation? How general are reactivity and selectivity differences between chiral Mo and Ru complexes? These were among several considerations that lead us to undertake the first set of studies involving a comparison of Mo and Ru-based chiral complexes for olefin metathesis. There are relatively few reports in the literature of direct comparisons between Mo and Ru complexes for olefin metathesis. Moreover, these reports generally relate to reactivity and not selectivity.

5.1a Ru-catalylzed Enantioselective Synthesis of Pyrans. Previously we reported on the Ru-catalyzed enantioselective synthesis of 2,4,6-substituted pyrans through the ring-opening metathesis reactions of meso oxabicycles.97 Since Ru complexes proved to be highly selective for this class of reactions this was to be the starting point of our comparative studies between Mo and Ru complexes. A representative example of this Ru-catalyzed process is shown in eq. 5.1 (see below). Exposure of oxabicycle 5.1 to 5 mol % of Ru complex 5.2 in the presence of two equivalents of styrene affords pyran 5.3 in 96% ee and 70% yield after 1 h. We refer to Ru complex 5.2 as a first generation chiral Ru catalyst. First generation Ru complexes possess a bidentate N-heterocyclic carbene ligand that is derived from a biaryl N,O-binapthyl intermediates. Another structural characteristic of these complexes is that they posses one alkoxy and one halide ligand, in contrast to the early achiral Ru complexes that typically have two halide ligands. (97) “Efficient Enantioselective Synthesis of Functionalized Tetrahydropyrans by Ru-Catalyzed Asymmetric Ring-Opening Metathesis/Cross-Metathesis (AROM/CM),” Gillingham, D. G.; Kataoka, O.; Garber, S. B.; Hoveyda, A. H. J. Am. Chem. Soc. 2004, 126, 12288−12290.

Chapter 5, Page 232

O5 mol % 5.2O

2 equiv. styreneC H , 1 h

NNCl

Ph

Me

MeMe

(5.1)

OBnOBn

RuO

Oi-Pr

5.3 96% ee, 70%6 6

5.1 After completing our studies into the Ru-catalyzed enantioselective synthesis of

pyrans our group developed a different class of chiral Ru complexes. In order to rigorously compare Mo and Ru complexes we first had to study the desymmetrization reactions of oxabicycles with our newly developed Ru catalysts 5.4 and 5.5 (eq. 5.2). These complexes, referred to as second generation chiral Ru complexes, are structurally very different from first generation Ru complexes. For one, the chirality of these complexes is derived from a C -symmetric chiral diamine backbone. Morevoer, the bidentate N-heterocyclic carbene ligands are derived from biphenyl intermediates, and not binapthyls as in the first generation complexes. An additional note about both classes of Ru complexes is that the halide ligand on the metal can be either a chloride (5.4) or an iodide (5.5). After understanding how both classes of Ru complexes behave in desymmetrization rea

2

ctions of oxabicycles we would be able to make a comparison with Mo complexes for this class of transformations.

O

OBn

5.3

OBn

Ostyrene

5.1

(5.2)

Ru

NNX

Ph

PhPh

or Mo complexes

5.5 X = I

Me

MeMe

5

O

Oi-Pr 5.4 X = Cl

.1b Ru-catalyzed Enantioselective Synthesis of Piperidines. The next phase of our comparative studies would entail studying Ru complexes for the desymmetrization of

n protected piperidines azabicycles. Of particular interest would be reactions where Mo complexes were found to be nonselective such as the enantioselective synthesis of N-B

Chapter 5, Page 233

(eq. 5.3, see below) where Mo complex 5.7 delivers piperidine 5.8 in only 21% ee. Other reactions of interest would include the desymmetrization reactions of carbamate-protected azabicycles. We expected the metathesis activity of Ru complexes would not be diminished by carbamate functional groups.

NBn

BnN

styreneC6H6,

22 C, 1 h 5.8 21% ee, 85%

Ru Catalysts??

OTBS OTBS

MeO

10 mol % 5.7(5.3)

5.6

5.2 Second Generation Ru Catalyst for the Enantioselective Synthesis of Pyrans

5.2a Comparison of Ru Catalysts for the Desymmetrization of Oxabicycles. During our investigations of Ru-catalyzed ring-opening/cross-metathesis reactions of oxabicycles we quickly took note of the reactivity differences between first and second generation Ru catalysts. Second generation Ru complexes were found to be m

10 equiv.

o

Mo

N

Ph

MeO

ore reactive than first generation complexes for this class transformations and this reactivity difference is exemplified in Scheme 5.1 (see below). Ru complex 5.10 catalyzes the ring-opening/cross-metathesis of oxabicycle 5.9 with styrene to provide pyran 5.11 in 94% ee and 50% yield after an extended reaction time of 36 h. This reaction must be carried out in the absence of solvent; if the reaction is performed with solvent <2% of the desired product is formed. We attribute this low reactivity to the lower strain energy of oxabicycle 5.9, in comparison to the desymmetrization of oxabicycle 5.1 that can be carried out in solvent (see eq. 5.1 above). Under identical reaction conditions, Ru complex 5.5 delivers the desired pyran in higher selectivity and significantly shorter reaction time. Treatment of oxabicycle 5.9 with 5 mol % of Ru complex 5.5 in twenty equivalents of styrene delivers pyran 5.11 in >98% ee and 55% yield.98 An additional note about these reactions is that iodide-containing Ru complexes were employed because in the reactions with chloride-containing complexes (5.2 in eq. 5.1 and 5.4 in eq. (98) The low yields in the formation of 5.11 are due to the volatility of the pyran.

Chapter 5, Page 234

5.2, see above), pyran 5.11 is formed in lower enantioselectivity.99 This trend in the reactivity of Ru complexes remained the same when we studied different oxabicyclic substrates. Although, it cannot be generalized that second generation catalysts are more selective than first generation catalysts for the enantioselective synthesis of pyrans.

OO

20 equiv. styreneno solvent,22 oC, 36 h

5.11 94% ee, 50%

5 mol % 5.10

Ru

N

O

N

I

Oi-Pr

Ph

5.9

PhPh

Scheme 5.1

Me

MeMe

MeNN

MeMe

OO

20 equiv. styreneno solvent,22 oC, 15 h

5.11 >98% ee, 55%

RuO

I

Oi-Pr

Ph 5 mol % 5.5

5.9 5.2b Comparison of First and Second Generation Ru Catalysts in the Preparation of Enantioenriched Pyrans. Illustrated in Figure 5.1 (see below) are the results we obtained in comparing Ru catalysts for the enantioselective synthesis of other pyran products. Ru complex 5.10 catalyzes the formation of iodide-containing pyran 5.12 in 93% ee and 65% yield after 4 h. In comparison, Ru complex 5.5 catalyzes the formation of pyran 5.12 in 82% ee and 70% yield after 1h. Here, the first generation Ru complex is the catalyst of choice delivering the desired pyran in greater selectivity. An advantage of

(99) The increase in selectivity for iodide-containing Ru complexes over chloride-containing complexes has been noted previously for a different class of chiral Ru complexes, see: “Enantioselective Ruthenium-Catalyzed Ring-Closing Metathesis,” Seiders, T. J.; Ward, D. W.; Grubbs, R. H. Org. Lett. 2001, 3, 3225−3228.

Chapter 5, Page 235

second generation Ru complexes over first generation complexes is that since reactions proceed at a faster rate, the temperature of certain reactions can be lowered to improve selectivity without having extended reaction times. As an example, Ru complex 5.10 furnishes pyran 5.13 in 90% ee and 85% yield after a reaction time of 36 h. Whereas Ru complex 5.5 provides pyran 5.13 in >98% ee and 84% yield after a reaction time of 15 h at –15 oC; at 22 oC pyran 5.13 is isolated in 92% ee. Clearly, performing this reaction at lower temperature with the first generation Ru complex would be impractical. Another example of the advantages in employing second generation Ru complexes, over first generation complexes, is in the preparation of fully substituted pyran 5.14. Ru complex 5.10 delivers pyran 5.14 in 80% ee and 44% yield after a long reaction time of 44 h. At a reaction temperature of –15 oC, second generation Ru complex 5.5 provides pyran 5.14 at 89% ee and 64% yield in 15 h; significantly shorter reaction time than with the first generation Ru complex. This pyran product is of special interest since it was employed by our group in the total synthesis of the natural product baconipyrone C.100

O

I

5.12 93% ee, 65%, 4 h2 mol % 5.10

5.12 82% ee, 70%, 1 h2 mol % 5.5

O

OBn

5.13 90% ee, 85%, 36 h2 mol % 5.10

5.13 >98% ee, 84%, 15 h2 mol % 5.5at −15 oC

O

OBn

5.14 80% ee, 44%, 44 h5 mol % 5.10

5.14 89% ee, 64%, 15 h2 mol % 5.5at −15 oC

MeMe

Figure 5.1. Comparison of Ru Complexes in theEnantioselective Synthesis of Pyrans

5.3 Mo Complexes for the Enantioselective Synthesis of Pyrans

5.3a Mo Complexes Prove to be Overly Reactive with Highly Strained Oxabicycles. The first oxabicyclic substrate we studied with chiral Mo complexes was oxabicycle 5.1 (eq. 5.4, see below), the strained oxabicycle that was readily desymmetrized by Ru complexes even in the presence of solvent. When performing a screen of Mo catalysts for this desymmetrization reaction we found that in all cases the starting oxabicycle underwent a ring-opening metathesis polymerization to deliver

(100) “Chiral N-Heterocyclic Carbenes in Natural Product Synthesis: Application of Ru-Catalyzed Asymmetric Ring-Opening/Cross-Metathesis and Cu-Catalyzed Allylic Alkylation to Total Synthesis of Baconipyrone C,” Gillingham, D. G.; Hoveyda, A. H. Angew. Chem., Int. Ed. 2007, 46, 3860−3864.

Chapter 5, Page 236

polymer 5.15. We performed this reaction under a variety of reaction conditions and were unable to prevent this polymerization from occurring.

O

OBn

Mo Catalysts

variousconditions

O

OBn n

5.15substrate polymerization

OBn

O

5.1

(5.4)

5.3b Mo-catalyzed Asymmetric Ring-Opening Cross-Metathesis of a Less Strained Oxabicycle. Since Mo complexes generally exhibit greater reactivity than Ru complexes, we turned our attention toward desymmetrization reactions of oxabicycles that Ru complexes could only perform in the absence of solvent. We postulated that less strained oxabicycles would not be polymerized by Mo complexes and w re ge ratified to find that this was indeed the case. The initial substrate under investigation was exocyclic-olefin containing oxabicycle 5.9. After a screen of available chiral Mo complexes we found the optimal catalyst to be adamantylimido alkylidene 5.7 and the metathesis with optimized reaction conditions is shown in eq. 5.5. Treatment of oxabicycle 5.9 with 5 mol % of Mo complex 5.7 in the presence of two equivalents of styrene delivers pyran 5.11 in 97% ee and 55% yield. The selectivity observed in this reaction was comparable with selectivities observed in reactions with Ru complexes (see Scheme 5.1 above), but due to the greater reactivity exhibited by Mo complexes the reaction was complete within significantly less reaction time (1 h with Mo complex 5.7 vs. 15 h and 36 h with Ru complexes).

OO

2 equiv. styreneC6H6, 22 oC

PhO

(5.5)

Mo

N

Me

Me

O

5 mol % 5.7

1 h

5.11 97% ee, 55%5.9

Chapter 5, Page 237

5.3c Mo-catalyzed Enantioselective Synthesis of Pyrans; Comparison of Reactions with Ru Complexes. Illustrated in Figure 5.2 are the results obtained when we studied the desymmetrization reactions of other less-strained oxabicycles with Mo complex 5.7. The optimal results with Ru complexes are also shown for comparison. As previously discussed, the first generation Ru complex 5.10 was the optimal catalyst for the formation of pyran 5.12, delivering the product in 93% ee. Mo complex 5.7 delivers pyran 5.12 in comparable selectivity of 96% ee, but within a slightly shorter reaction time of 1 h vs. 4 h with Ru complex 5.10. This is a case in which chiral Mo and Ru complexes exhibit similar reactivity and selectivity profiles. Next, we compared the enantioselective synthesis of pyran 5.13, in which the second generation Ru catalyst 5.5 was the optimal for the reaction delivering the desired pyran in >98% ee within 15 h at –15 oC. Mo complex 5.7 is outperformed by Ru complex 5.5 in this particular reaction. Mo complex 5.7 delivers pyran 5.13 in 90% ee and 81% yield within 1 h at 22 oC. Clearly, the Ru catalyst would be the catalyst of choice in this transformation given the greater selectivity, despite the reaction being complete in shorter reaction time with the Mo complex. When we compared the enantioselective synthesis of fully-substituted pyran 5.14 we found again that the Ru complex outperformed the Mo complex. Second generation Ru complex 5.5 provides pyran 5.14 in 89% ee within 15 h at –15 oC. Mo complex 5.7 delivers pyran 5.14 within shorter reaction time at 22 oC (6 h vs. 15 h), however, the product is isolated in only 67% ee (vs. 89% ee with Ru complex 5.5). These studies into the enantioselective synthesis of pyrans were key in providing us with further insight into the intrinsic differences and similarities between Ru and Mo complexes.

I

O

5.12 93% ee, 65%, 4 h2 mol % 5.10

5.12 96% ee, 76%, 1 h2 mol % 5.7

O

OBn

5.13 >98% ee, 84%, 15 h2 mol % 5.5at −15 oC

5.13 90% ee, 81%, 1 h5 mol % 5.7

O

OBn

5.14 89% ee, 64%, 15 h5 mol % 5.5at −15 oC

5.14 67% ee, 70%, 6 h5 mol % 5.7

MeMe

Figure 5.2. Comparison of Mo and Ru Complexesfor the Enantioselective Synthesis of Pyrans

Chapter 5, Page 238

Most distinguishing, these investigations revealed for the first time that Mo and Ru complexes complement each other for the same set of enantioselective olefin metathesis transformations. Before concluding this section it should be noted that in all cases the asymmetric reactions with Mo and Ru complexes delivered the opposite antipodes of pyran products.101 To avoid confusion and for clarity of comparison enantioselectivites are always given as positive.

5.4 Ru Complexes for the Enantioselective Synthesis of Carbamate-protected Piperidines

5.4a Ru Complexes Prove Inefficient for the Synthesis of 2,4,6-substituted Carbamate-protected Piperidines. After comparing Mo and Ru complexes for the enantioselective synthesis of pyrans, we turned our focus to comparing these complexes for the enantioselective synthesis of piperidines. As previously noted, Mo complexes proved quite efficient and selective for the synthesis of various enantioenriched piperidines. The first desymmetrization substrate we studied with Ru complexes was Cbz-protected azabicycle 5.16 (eq. 5.6). Screening of available chiral complexes under a variety of reaction conditions revealed that this substrate was incompatible with Ru complexes for the asymmetric ring-opening/cross-metathesis with styrene; in all cases azabicycle 5.16 was isolated in >98% yield. Furthermore, the diastereomer of 5.16, which contains the opposite stereochemistry at the hydroxy group carbon was also found to be incompatible with Ru complexes.

OTBS Screen ofRu CatalystsCbz

N variousconditions

<2% conv.

5.16

(5.6)

5.4b Ru Complexes Prove to be Selective for the Synthesis of 2,6-substituted Carbamate-protected Piperidines. Undaunted by the initial lack of reactivity of Ru complexes with azabicyclic substrates, we explored the desymmetrization of other carbamate-protected azabicycles. We discovered that Ru complexes are compatible with azabicycles that deliver 2,6-substituted piperidines and results are shown in Scheme 5.2 (see below). To this end, first generation Ru complex 5.2 was the optimal catalyst for the desymmetrization of azabicycles 5.17 and 5.19. Treatment of azabicycle 5.17 with 10 mol % Ru catalyst 5.2 in 20 equivalents of styrene delivers piperidine 5.18 in 90% ee and 70% yield after 24 h. And treatment of azabicycle 5.19 with 10 mol % Ru catalyst 5.2 in

(101) See the experimental section of this chapter for HPLC traces of pyran products.

Chapter 5, Page 239

20 equ

ol % for the synthesis of pyrans

trization reactions employing Mo complexes (24 h). In both of these

ivalents of styrene furnishes piperidine 5.20 in 94% ee and 66% yield. Several aspects of these reactions, in comparison with reactions Mo complexes, merit mention. For one, similar to the desymmetrization reactions of less strained oxabicycles, reactions with Ru complexes require that they be performed in the absence of solvent. Additionally, these reactions with Ru complexes require a catalyst loading of 10 mol %, in comparison with Ru catalyst loadings as low as 2 m(Fig. 5.1, see above). The reaction times obtained in these reactions are comparable with the desymmereactions Ru complex 5.2 delivers the product in greater enantioselectivity than the optimal Mo complex (82% ee for 5.18 and 90% ee for 5.20, see Ch 3 Fig 3.4 above). An advantage the optimal Mo complex has over the optimal Ru complex for these reactions is that with the Mo complex a catalyst loading of 5 mol % is sufficient for the reaction to proceed, as opposed to a catalyst loading of 10 mol % with the Ru complex.

N10 mol % 5.2

N

O

N

Cl

Oi-Pr

Ph

Scheme 5.2

Me

MeMe

CbzNCbz

5.18 90% ee, 70%

20 equiv.styrene

no solvent22 oC, 24 h

Ru

10 mol % 5.2O

OEt

5.17

5.19

NCO2Et

5.20 94% ee, 66%

20 equiv.styrene

no solvent22 oC, 24 h

N

5.4c Mechanistic Rationale for the Difference in Reactivity of Ru Complexes with Different Carbamate Protected Azabicycles. In order to understand the reasons for differences in the reactivity of Ru complexes toward different azabicycles we conducted energy minimization calculations of the azabicyclic substrates.102 Our analysis indicates that the conformations of hydroxy-substituted carbamates 5.16 and 5.21 prevent Ru

(102) Energy minimization calculations were performed on Spartan version 5.0.

Chapter 5, Page 240

complexes from reacting with the norbornyl olefins because of distortions of the [3.2.1]-azabicycle caused by the large tert-butyldimethylsilyl ethers (Fig. 5.3, see below). In line with previous studies involving ring-opening/cross-metathesis reactions of norbornenes and oxabicycles it is likely that the initial approach of a Ru catalyst toward azabicyclic olefins is from the exocyclic face. This is shown for the azabicycle 5.17, which is compatible with Ru complexes. Now, in the case of azabicycle 5.16, the bulky tert-butyldimethylsilyl group causes an increase in the bond angle shown bringing the carbamate group on the nitrogen to be in closer proximity to the azabicyclic olefin. This most likely prevents Ru complexes from forming the metallacyclobutane necessary for the initial ring-opening metathesis of the substrate. This effect is even more prominent in the case of azabicycle 5.21, where a syn-pentane interaction is present between the tert-butyldimethylsilyl group and the olefinic carbons. This steric repulsion causes the olefinic carbons to be in closer proximity to the carbamate group; just as in azabicycle 5.16. The difficulties encountered in trying to desymmetrize 5.16 and 5.21 underline the need to develop new chiral Ru complexes, either smaller or more reactive, that can overcome these metathesis inhibiting effects

NOBn

OSiO H

N

OBn

OH

H

5.16 5.215.17

Figure 5.3. Rationale for Reactivity of Ru Complexes Toward Azabicycles

Approach ofRu Catalyst

MeMe

Me MeMe

moreproximal

H

increased bond angle

NOBn

OH

moreproximal

OTBS

increased bond angle

5.5 Ru Complexes for the Enantioselective Synthesis of N-alkyl Piperidines

5.5a Ru-catalyzed Asymmetric Ring-Opening/Cross-Metathesi of N-Me Azabicycles. Our next efforts involved the Ru-catalyzed desymmetrization reaction of N-Me-containing azabicycle 5.22 (eq. 5.7).

Chapter 5, Page 241

NMe20 equiv.

styreneno solvent22 oC, 36 h

MeN

OTBS

5.23 30% ee, 33% conv.

5 mol % 5.2

Ru

N

O

N

Oi-Pr

Ph

5.22

Me

MeMe

Previously, we demonstrated that this substrate was readily and efficiently

desymmetrized by Mo complexes. Ru complexes, on the other hand, were found to be inefficient at performing this transformation. Treatment of azabicycle 5.22 with 5 mol % of Ru catalyst 5.2 in 20 equivalents of styrene delivers piperidine 5.23 in only 33% ee and 30% conversion to desired product after a reaction time of 36 h. This result was not surprising since earlier studies revealed Ru complexes to be incompatible with amine-containing ring-closing metathesis substrates.

OTBS

Cl

(5.7)

5.5b Ru-catalyzed Asymmetric Ring-Opening/Cross-Metathesis of N-Bn Azabicycles. The next azabicyclic substrate we studied with Ru complexes was N-Bn-containing azabicycle 5.24 (eq. 5.8). These investigations resulted in one of the most intriguing differences we had yet to observe between Mo and Ru complexes. Previously we found that Mo complexes catalyzed the desymmetrization of 5.24 in low enantioselectivity and to our surprise we found Ru complexes were more selective at performing this transformation. Toward this end, treatment of N-Bn azabicycle 5.24 with 5 mol % of Ru complex 5.2 in twenty equivalents of styrene delivers desired piperidine

-Me azabicy

to prevent the nitrogen from binding to Ru, as strongly as in the case of the N-Me, and thus allows the catalyst to perform the reaction.

5.25 in >98% ee and 80% yield after 24 h. It is likely that in the case of N

cles, the nitrogen binds to the Ru metal center and inhibits metathesis activity; this is evident from the low conversion to desired product (see eq. 5.7 above). The benzyl group is most likely large enough

Chapter 5, Page 242

NBn20 equiv.

styreneno solvent22 oC, 24 h

OTBS

BnN

OTBS

5.25 >98% ee, 80%

5 mol % 5.2

Ru

N

O

NCl

Oi-Pr

Ph

5.24

(5.8)

Me

MeMe

e 5.3, see below). reatm

5.5c Ru-catalyzed Desymmetrization Reactions of Alcohol-containing Azabicycles. The next logical step for us to take was to study the desymmetrization reaction of N-Bn azabicycles that contained secondary alcohols, in place of previously studied silyl-protected alcohols. As mentioned earlier, Ru complexes unlike Mo complexes are tolerant of alcohol functionalities. To this end, we studied the desymmetrization reaction of azabicycle 5.26 with first gerneration and second generation chiral Ru catalysts, 5.2 and 5.4, respectively (SchemT ent of N-Bn azabicycle 5.26 with 5 mol % of first generation Ru complex 5.2 in twenty equivalents of styrene delivers the desired piperidine 5.27 in >98% ee, >98% conversion to desired product, and 82% yield after 24 h. Treatment of N-Bn azabicycle 5.26 with 5 mol % of second generation Ru complex 5.4 in twenty equivalents of styrene delivers desired piperidine 5.27 in 94% ee, 80% conversion to desired product, and 65% yield after 24 h. As displayed by these results, the first generation Ru complex is the optimal catalyst delivering the product with greater efficiency, albeit both complexes deliver the desired product in comparable levels of selectivity. This distinction in reactivity difference between the first and second generation Ru complexes holds true in desymmetrization reactions of azabicycle 5.26 with other aryl olefin cross-partners.

Chapter 5, Page 243

BnN

NBn20 equiv.

styreneno solvent

OH OH

10 mol % 5.2

Ru

N

O

NCl

Oi-Pr

Ph

Scheme 5.3

Me

MeMe

5.27 >98% ee, 82%>98% conv.22 oC, 24 h

BnN

NBn

5.27 94% ee, 65%80% conv.

20 equiv.styrene

no solvent22 oC, 24 h

OH OH

Ru

N

O

NCl

Oi-Pr

Ph

PhPh

5 mol % 5.4

5.26

5.26

Me

MeMe

5.5d Comparison of Ru Complexes in the Desymmetrization of Alcohol-containing

Azabicycles with Various Aryl Olefin Cross-Partners. Illustrated in Figure 5.3 (see below) are results obtained when we performed the desymmetrization of azabicycle 5.26 with electronically and sterically modified styrene cross-partners. First we studied the desymmetrization reaction with o-Me-styrene. Ru complex 5.2 delivers piperidine 5.28 in >98% ee and 81% yield. Ru complex 5.4 provides piperidine 5.28 in >98% ee and 60% yield. Next, we studied the desymmetrization reaction of 5.26 with electron rich p-OMe-styrene. Ru complex 5.2 delivers piperidine 5.29 in >98% ee and 67% yield. Ru complex 5.4 provides piperidine 5.29 in 96% ee and 38% yield. Last, we studied the desymmetrization reaction of 5.26 with electron poor p-CF3-styrene. Ru complex 5.2 delivers piperidine 5.30 in 97% ee and 78% yield. Ru complex 5.4 provides piperidine 5.30 in 96% ee and 63% yield. It is clear from these results that the first generation Ru complex 5.2 is the catalyst of choice for the desymmetrization of N-Bn-containing

Chapter 5, Page 244

azabicycles. It should be noted that these desymmetrization reactions were performed with the chloride-containing Ru complexes since the enantioenriched piperidines were isolated in high selectivity.

NBn

OH

5.28 >98% ee, 81%>98% conv.5 mol% 5.2

5.28 >98% ee, 60%95% conv5 mol % 5.4

5.29 >98% ee, 67%87% conv.5 mol% 5.2

5.29 96% ee, 38%50% conv.5 mol% 5.4

Me

NBn

OH

OMe

5.30 97% ee, 78%>98% conv.5 mol% 5.2

5.30 96% ee, 63%85% conv.5 mol% 5.4

NBn

OH

CF3

Figure 5.3. Scope of AROM/CM with Various Cross-Partners 5.6 Conclusions

This chapter detailed the first direct comparison of chiral Mo and Ru complexes for asymmetric olefin metathesis. Our studies clearly indicate the importance of having families of both classes of chiral complexes due to their complementarity in reactivity and selectivity. Ru complexes are the preferred complexes for the desymmetrization of highly strained oxabicycles given that Mo complexes tend to polymerize these type of substrates. When it comes to the desymmetrization of less strained oxabicycles, the nature of the substrate dictates whether a Mo or Ru complex will be the optimal catalyst. A pronounced distinction between these two classes of complexes was observed in the desymmetrization reactions of azabicycles. Mo complexes outperformed Ru complexes for the enantioselective synthesis of 2,4,6-substituted carbamate-protected piperidines. Whereas Ru complexes proved to deliver 2,6-substituted carbamate-protected piperidines in comparable levels of selectivity. The most striking difference between Mo and Ru complexes was in the desymmetrization reactions of N-alkyl azabicycles. Mo complexes efficiently and selectively deliver a variety of N-Me-containing piperidine products. In contrast, Ru complexes appear to be incompatible with N-Me azabicycles, delivering products in low selectivity and efficiency. The ability of each complex to selectively furnish N-Bn azabicycles was reversed from that of N-Me azabicycles. Ru complexes delivered products in higher enantioselectivity than their Mo counterparts. In conclusion, these studies are of utmost importance in our continuing pursuit to understand the

Chapter 5, Page 245

fundamental intricacies of Mo and Ru complexes, as well as guiding us toward future endeavors.

5.7 Experimental Section103

Characterization data for pyran products: All oxabicycles were prepared according to published literature procedures and characterization data for all pyrans was previously reported.104

Pyran 5.11. The enantiomeric purity of this compound was determined by HPLC analysis in comparison with authentic racemic material, shown below: Chiralpak OB-H (4.6 x 250 mm), 99.8:0.2 hexanes:i-PrOH, 0.5 mL/min, λ = 254 nm.

Authentic racemic >98% ee with Ru catalyst 5.5

O

5.11

Authentic racemic 97% ee with Mo catalyst 5.7

determined by variable temperature 1H NMR analysis. Additionally, this is likely the reason that in some instances 13C NMR spectra do not appear to have the correct number of carbon peaks. In these instances the 13C NMR spectra is reported as observed. (104) For HPLC traces of pyrans with Ru complex 5.2, the trace of pyran 5.12 with Ru complex 5.10, and the trace of pyran 5.14 with Ru complex 5.5, see supporting information in: “Efficient Enantioselective Synthesis of Functionalized Tetrahydropyrans by Ru-Catalyzed Asymmetric Ring-Opening

(103) In some instances, rotamers of carbamate- and amide-containing azabicycles were observed as

Metathesis/Cross-Metathesis (AROM/CM),” Gillingham, D. G.; Kataoka, O.; Garber, S. B.; Hoveyda, A. H. J. Am. Chem. Soc. 2004, 126, 12288−12290.

Chapter 5, Page 246

Pyran 5.12. The enantiomeric purity of this compound was determined by HPLC analysis in comparison with authentic racemic material, shown below: Chiralpak OJ (4.6

-PrOH, 1.0 mL/min, λ = 254 nm.

x 250 mm), 99.8:0.2 hexanes:i

O

82% ee with Ru catalyst 5.5 Authentic racemic

5.12

I

Authentic racemic 97% ee with Mo catalyst 5.7

Pyran 5.13.105 The enantiomeric purity of this compound was determined by HPLC analysis in comparison with authentic racemic material, shown below: Chiralpak OJ (4.6 x 250 mm), 99.5:0.5 hexanes:i-PrOH, 1.0 mL/min, λ = 254 nm.

O

5.13

OBn

(105) Two sets of conditions for HPLC separation of enantiomers are shown because the data was acquired at different times.

Chapter 5, Page 247

Authentic racemic >98% ee with Ru catalyst 5.5 at −15 oC

Chiralpak AD (4.6 x 250 m ), 99.9:0.1 hexanes:i-PrOH, 0.2 mL/min, λ = 254 nm. mAuthentic racemic 90% ee with Mo catalyst 5.7

Pyran 5.14. The enantiomeric purity of this compound was determined by HPLC analysis in comparison with authentic racemic material, shown below: Chiralpak AD (4.6 x 250 mm), 96:4 hexanes:i-PrOH, 0.8 mL/min, λ = 210 nm. Authentic racemic 67% ee with Mo catalyst 5.7

Synthesis of Azabicycles: The synthesis of N-Me azabicycle 5.22 was reported earlier.106 The synthetic route used to form 5.22 was modified and employed for the (106) See the experimental section of Ch 3.

O

OBn

5.14

MeMe

Chapter 5, Page 248

preparation of azabicycle 5.26, shown in Scheme 5.4 (see below). Alkylation of p-methoxypyridine with vinylmagnesium bromide in the presence of benzoyl chloride provides piperidinone 5.31. Conjugate addition to 5.31 with a mixed vinyl hight-order cuprate reagent then delivers divinyl piperidinone 5.32. Then, ring-closing metathesis of 5.32 with Ru catalyst 5.33107 furnishes azabicycle 5.34. Selective reduction of the carbonyl in 5.34 with L-Selectride® followed by reduction of the benzamide to the tertiary benzyl amine ultimately furnishes metathesis substrate 5.26.

Scheme 5.4. Synthesis of Azabicycle 5.26

N

OMe

N

O

MgBr, BzCl

THF−78 oC to −15 oC

1 h

MgBrCuCN, MeLi,

CH2Cl2, 22 oC18 h

1) L-Selectride®, THF,− o o

O

N

OH

5.31 96% yield

THF−78 oC to −15 oC

7 h

N

O

Bz

5.32 90% yield

two steps

Ru

NAr

Cl

ArNCl

Oi-Pr5 mol % 5.33

Ar = 2,4,6-Me C H

Bz

78 C to 22 C, 1 h

2) DIBAL-H, THF,−78 oC to 22 oC, 3 h

BzN Bn

5.34 60% yield5.26

40% yield over

3 6 2

Synthesis of roun bottoman addition funnel, was charged with p-methoxypyridine (5.00 g, 46.2 mmol), THF (46 mL), and the reaction vessel was allowed to cool to −78 °C in a dry ice/acetone bath. To this solution was added vinylmagnesium bromide (55.5 mL, 55.5 mmol, 1.0 M in THF) dropwise over a period of 10

min, after which time the resulting mixture was allowed to warm to –15 °C in a cooling bath. To this mixture was added benzoyl chloride (8.48 mL, 69.4 mmol) dropwise over a period of 10 min. The mixture was allowed to stir at –15 °C for 20 min, after which time the flask was removed from the cooling bath and the reaction was quenched by the addition of a 2.0 M aqueous solution of HCl (75 mL). The resulting solution was allowed to stir at 22 °C for 15 min, after which point it was washed with Et2O (4 × 100 mL). The

Piperidinone 5.31. A 250-mL d- flask, fitted with N

O

Bz

5.31

combined organic layers were washed with a 1.0 M aqueous solution of NaOH (25 mL),

(107) “Efficient and Recyclable Monomeric and Dendritic Ru-Based Metathesis Catalysts,” Garber, S. B.;

. J. Am. Chem. Soc. 2000, 122, 8168−8179. Kingsbury, J. S.; Gray, B. L.; Hoveyda, A. H

Chapter 5, Page 249

a saturated aqueous solution of NaCl (50 mL), dried (MgSO4), filtered, and the volatiles removed in vacuo. The resulting dark brown residue was purified by silica gel chromatography (dry load, 1:1 to 3:1 Et2O:hexanes) to give 5.31 as slight yellow solid (10.05 g, 44.20 mmol, 96%). mp: 40−43 oC. IR (neat): 3081 (w), 2995 (w), 1666 (s), 1598 (s), 1393 (m), 1331 (s), 1288 (s), 1213 (s), 1145 (s). 1H NMR (400 MHz, CDCl3, TMS): δ 7.56−7.45 (m, 6H), 5.87 (ddd, J = 17.2, 10.6, 4.9 Hz, 1H), 5.44−5.40 (m, 1H), 5.30−5.27 (m, 2H), 5.21 (dd, J = 17.2, 1.6 Hz, 1H), 2.98 (dd, J = 16.7, 6.6 Hz, 1H), 2.68 (d, J = 16.7 Hz, 1H). 13C NMR (100 MHz, CDCl3): δ 192.6, 172.0, 142.8, 132.9, 132.7, 131.8, 128.9, 128.5, 118.0, 107.8, 54.4, 40.2. HRMS ES (m/z) Calcd for C14H14NO2 228.1025 (M + H)+, Found 228.1024.

Synthesis of cis-2-vinylpiperidine 5.32. A 500-mL round bottom flask, fitted with an addition funnel, was charged with CuCN (2.96 g, 33.1 mmol), THF (51 mL), and the reaction vessel was allowed to cool to −78 °C in a dry ice/acetone bath. To this solution was added MeLi (21.4 mL, 33.1 mmol, 1.54 M in Et2O) dropwise over a period of 10

min. At this time, the reaction flask was removed from the dry ice bath and placed in an ice bath at 0 °C for 1 min, after which time the flask was allowed to cool to –78 °C in a dry ice/acetone bath. To this mixture was added vinylmagnesium bromide (34.0 mL, 33.1 mmol, 0.97 M) in THF, dropwise over a period of 15 min and the mixture was allowed to stir at –78 °C for 10 min. The mixture was then charged with a solution of 5.31 (5.00 g, 22.0 mmol, as a solution in 20 mL) of THF in a dropwise fashion over a period of 40 min, and the resulting orange mixture was allowed to stir at –78 °C for 5 h (it is critical to add 5.31 slowly to prevent formation of the trans product). At this time, the mixture was poured into a solution consisting of a saturated aqueous solution of NH4Cl and a saturated aqueous solution of NH4OH (9:1, 250 mL)108 at 22 oC. The mixture was allowed to stir for 12 h, after which time the solution was washed with EtOAc (3 × 250 mL). The combined organic layers were washed with a saturated aqueous solution of NaCl (150 mL), dried (MgSO4), filtered, and the volatiles removed in vacuo. The resulting dark brown residue was purified by silica gel chromatography (dry load, 4:1 to 1:1 hexanes:EtOAc) to provide 5.32 as colorless oil (5.05 g, 19.8 mmol, 90%). IR (neat): 3081 (w), 2974 (w), 2905 (w), 1722 (s), 1634 (s), 1394 (s), 1338 (s), 1224 (m), 916 (m), 702 (m). 1H NMR (400 MHz, CDCl3, TMS): δ 7.48−7.41 (m, 5H), 5.98 (ddd, J = 17.2, 10.6, 5.1 Hz, 2H), 5.23−5.18 (m, 6H), 2.69 (d, J = 5.7 Hz, 4H). 13C NMR (100 MHz, CDCl3): δ 206.1, 172.0, 138.2, 136.1, 130.2, 128.9, 128.7, 126.6, 117.2,

N

O

Bz

5.32

(108) Caution should be taken as this part of the procedure is exothermic.

Chapter 5, Page 250

54.7, 43.2. HRMS ES (m/z) mass Calcd for C16H18NO2 256.1338 (M+H)+, Found 256.1340.

Synthesis of Azabicycle 5.34. A 100-mL round-bottom flask was charged with 5.32 (610 mg, 2.39 mmol), CH2Cl2 (25 mL), and Ru catalyst 5.33 (75 mg, 0.12 mmol, 5 mol %) at 22 °C. The reaction mixture was allowed to stir for 18 h, after which time the volatiles were removed in vacuo. The resulting dark brown

residue was purified by silica gel chromatography (dry load, 4:1 to 1:1 hexanes:EtOAc) to give 5.34 (323 mg, 1.42 mmol, 60%) as slight brown oil. IR (neat): 3057 (w), 2976 (w), 2901 (w), 1716 (s), 1635 (s), 1412 (s). 1H NMR (400 MHz, CDCl3, TMS): reported as a 1:1 mixture of amide rotamers, δ 7

BzN

O

5.34

.55−7.42 (m, 5H), 6.33 (br s, 1H), 6.21 (br s, 1H), 5.35 (br s, 1H), 41H). 13C NMR (127.3, 58.7, 55.0(M+H)+, Found 2

Synthesi

)

.74 (br s, 1H), 3.05−2.95 (m, 1H), 2.58−2.44 (m, 2H), 2.38−2.28 (m, 100 MHz, CDCl3): δ 205.1, 168.1, 134.8, 134.6, 133.4, 130.8, 128.8, , 46.7, 45.6. HRMS ES (m/z) mass Calcd for C14H14NO2 228.1025 28.1018.

s of Azabicyclic Precursor to 5.26 (shown here on the left).109 A 200-mL round-bottom flask was charged with 5.34 (1.35 g, 5.94 mmol), THF (59 mL), and the reaction vessel was allowed to cool to −78 °C in a dry ice/acetone bath. To this solution was added L-Selectride® (6.53 mL, 6.53 mmol, 1.00 M in

THF) dropwise over a period of 10 min, after which time the flask was removed from the dry ice batch and the mixture was allowed to stir at 22 °C for 1 h. The reaction vessel was allowed to cool to 0 °C in an ice-bath and the reaction was quenched by the sequential slow addition of a 1.0 M solution of NaOH (20 mL) and H2O2 (37 wt % in H2O, 20 mL). At this point, the flask was removed from the ice-bath and the mixture was allowed to stir at 22 °C for 15 min, after which time a 1.0 M solution of HCl (20 mL) was added. The resulting solution was washed with CH2Cl2 (3 × 100 mL . The combined organic layers were dried (MgSO4), filtered, and the volatiles removed in vacuo. The remaining yellow residue was purified by silica gel chromatography (dry load, 2:1 to 1:2 hexanes:EtOAc) to deliver the precursor to 5.26 as colorless oil (1.09 g, 4.75 mmol, 80%). IR (neat): 3408 (br), 2974 (w), 2945 (m), 2943 (m), 1621 (s) 1451 (s) 1067 (s), 696 (s). 1H NMR (400 MHz, CDCl3, TMS): δ 7.50−7.39 (m, 5H), 6.51 (dd, J = 5.9, 2.0 Hz, 1H), 6.36 (dd, J = 5.9, 2.0 Hz, 1H), 5.14 (br s, 1H), 4.47 (br s, 1H), 4.05−4.01 (m, 1H), 2.50−2.44 (m, 1H), 2.24 (d, J = 9.7 Hz, 1H), 2.14−2.08 (m, 1H), 1.91 (d, J = 14.8

BzN

OH

(109) The precursor to 5.26 is the azabicycle containing a secondary alcohol derived from hydride reduction of the carbonyl in 5.34. The diastereomeric identity of this alcohol (endo vs. exo) was assigned

n of Ch 3. by correlation to related azabicyclic substrates. See the experimental sectio

Chapter 5, Page 251

Hz, 1H), 1.79 (d, J = 14.8 Hz, 1H). 13C NMR (100 MHz, CDCl3): δ 167.3, 136.7, 135.6, 133.2, 130.3, 128.6, 127.3, 65.7, 60.1, 55.8, 37.3, 35.5. HRMS ES (m/z) mass Calcd for

e it was washed with EtOAc (3 × 100 mL). The combinevacuo. T94:5:1 C50%). m

C14H15N

BnN

OHO2 229.1098 (M )+, Found 229.1102.

Synthesis of Olefin Methathesis Substrate 5.26. A 200-mL round-bottom flask, fitted with an addition funnel, was charged with the azabicyclic precursor to 5.26 (1.09 g, 4.75 mmol), THF (48 mL), and the mixture was allowed to cool to −78 °C in a dry ice/acetone bath. To this solution was added DIBAL-H (16.9 mL, 23.8 mmol, 20% by volume in hexane) dropwise over a period of 15 min.

The resulting mixture was allowed to stir at –78 °C for 10 min, after which time the flask was removed from the dry ice bath and the mixture was allowed to stir at 22 °C for 2 h. At this time, the mixture was allowed to cool to 0 °C in an ice-bath and the reaction was quenched by the addition of a saturated aqueous solution of potassium sodium tartrate (20 mL). At this point, the flask was removed from the ice-bath and the mixture was allowed to stir at 22 °C for 15 h, after which tim

5.26

d organic layers were dried (Na2SO4), filtered, and the volatiles removed in he dark brown residue was purified by silica gel chromatography (CH2Cl2 to

H2Cl2:MeOH:NEt3) to provide 5.26 as slight yellow solid (510 mg, 2.37 mmol, p: 102−104 oC. IR (neat): 3408 (br), 2930 (s), 2836 (w), 1489 (m) 1438 (m),

1042 (s). 1H NMR (400 MHz, CDCl3, TMS): δ 7.26−7.21 (m, 5H), 6.29 (s, 2H), 3.90−3.85 (m, 1H), 3.53 (s, 2H), 3.50 (br s, 2H), 2.28 (br s, 1H), 2.24 (dd, J = 6.0, 3.7 Hz, 1H), 2.20 (dd, J = 6.0, 3.7 Hz, 1H), 1.81−1.80 (m, 1H), 1.78−1.77 (m, 1H). 13C NMR (100 MHz, CDCl3): δ 139.9, 135.3, 128.6, 128.3, 126.8, 66.1, 64.0, 57.8, 37.8. HRMS ES (m/z) mass Calcd for C14H18NO 216.1388 (M+H)+, Found 216.1399.

Synthesis of Olefin Metathesis Substrate 5.24.110 Azabicycle 5.24 was prepared by TBS protection of the alcohol in 5.26. IR (neat): 2925 (s), 2853 (w), 1252 (w), 1065 (s). 1H NMR (400 MHz, CDCl3): δ 7.38 (d, J = 7.2 Hz, 2H), 7.31 (t, J = 7.2 Hz, 1H), 7.24 (d, J = 7.2 Hz, 2H), 5.98 (br s, 2H), 4.00

(dd, J = 5.2 Hz, 1H), 3.60 (br s, 2H), 3.46 (br s, 2H), 2.14 (br s, 2H), 1.56 (d, J = 12.0, 2H), 0.85 (s, 9H), −0.44 (s, 6H). 13C NMR (100 MHz, CDCl3): δ 133.7, 132.6, 128.8, 126.8, 65.2, 64.1, 57.7, 37.5, 25.9, 17.9, 0.02, −4.7. HRMS ES (m/z) mass Calcd for C20H31NOSi 329.2175 (M)+, Found 329.2184.

BnN

OTBS

5.24

(110) For the TBS protection of a related azabicycle see the experimental section of Ch 3.

Chapter 5, Page 252

Representative procedure for Ru-catalyzed asymmetric ring-opening metathesis/cross-metathesis reactions of azabicycles, synthesis of piperidine 5.8: Ru

complex 5.2 (3.8 mg, 0.0046 mOTBS mol) was dissolved in styrene

volatiles

), 5.98 (dd, J = 15.9, 8.6 Hz, 1H), 5.82 (ddd, J = 17.2, 10.1, 8.4 Hz, 1H), 5.21 (dd, J = 17.2, 1.5 Hz, 1H), 5.07 (dd 10.1, 1.5 Hz, 1H), 3.94

(d, J = 15.5 Hz, 1H), 3.74 (d, J = 15.5 Hz, 1H), 3.76 3.68 (m, 1H), 3.11 (ddd, J = 11.6, 8.8, 2.8 Hz, 1H), 3.05 (ddd, J = 11.2, 8.4, 2.8 Hz, 1H), 1.96 1.90 (m, 2H), 1.64 (br s, 1H), 1.58−1.48 (m, 2H). C NMR (100 MHz, CDCl ): 142.4, 140.1, 137.1, 134.2, 130.2, 129.2, 128.5, 127.9, 127.4, 126.3, 126.2, 115.9, 68.0, 64.4, 64.1, 55.2, 43.0. HRMS EI (m/z) mass Calcd for C H NO 319.1936 (M) , Found 319.1938. The enantiomeric purity of this compound was determined by HPLC analysis in comparison with authentic racemic m n below: Chiralpak OD (4.6 x 250 mm), 98:2 hexanes:i-PrOH, 1.0 mL/min, = 254 nm.

(106 µL, 0.930 mmol) in a 4-mL vial. The resulting mixture was added by syringe to a solution of azabicycle 5.6 (20.0 mg, 0.0930 mmol) in styrene (106 µL, 0.930 mmol) in a 4-mL vial. The mixture was allowed to stir for 24 h, upon which time the

were removed in vacuo. The resulting dark brown residue was purified by silica gel column chromatography (98:2 CH2Cl2:MeOH) to give piperidine 5.8 as colorless oil (23.8 mg, 0.0744 mmol, 80%). IR (neat): 3363 (s), 2955 (m), 2930 (m), 2338 (s), 1104 (s). 1H NMR (400 MHz, CDCl3): δ 7.29−7.12 (m, 10H), 6.44 (d, J = 15.9 Hz, 1H), 5.99 (dd, J = 15.9, 8.6 Hz, 1H), 5.81 (ddd, J = 17.6, 10.0, 8.8 Hz, 1H), 5.18 (dd, J = 17.2, 0.9 Hz, 1H), 5.04 (dd, J = 10.2, 1.5 Hz, 1H), 3.92 (d, J = 15.3 Hz, 1H), 3.73 (d, J = 15.3 Hz, 1H), 3.69−3.64 (m, 1H), 3.07 (ddd, J = 14.4, 9.2, 2.8 Hz, 1H), 3.01 (ddd, J = 11.2, 8.8, 2.4 Hz, 1H), 1.80−1.76 (m, 2H), 1.62−1.51 (m, 2H), 0.86 (s, 9H), 0.04 (s, 6H). 13C NMR (100 MHz, CDCl3): δ. 142.7, 140.1, 137.2, 134.6, 130.0, 129.3, 128.5, 127.8, 127.3, 126.3, 126.2, 115.6, 68.6, 64.4, 64.0, 55.2, 43.4, 43.4, 25.9, 18.2, −4.5. HRMS EI (m/z) mass Calcd for C28H40NOSi 434.2879 (M+H)+, Found 434.2889. [α]D

20 −63.52 (c = 0.5, CHCl3) for a sample of 94% ee. Determination of enantiomeric excess of 5.25 involved piperidine 5.27, derived from TBAF deprotection of the secondary alcohol in 5.25.

Piperidine 5.27. IR (neat): 3333 (br), 3024 (w), 2936 (m), 2911 (m), 1495 (m), 1457 (m). 1H NMR (400 MHz, CDCl3, TMS): δ 7.30−7.09 (m, 10H), 6.45 (d, J = 15.9 Hz, 1H

n

OH

NB5.8

NBn

5.27 , J =−

−13

3 δ

22 25+

aterial, showλ

Chapter 5, Page 253

Authentic racemic >98% ee with Ru catalyst 5.2

Authentic racemic ee with Ru catalyst 5.4 94%

r piperidine products:

Piperidine 5.28. IR (neat): 3333 (br), 2930 (w), 2917 (w), 2836

Characterization data fo

1

13

+

25 he nantiomeric purity of this compound was determined by HPLC analysis in comparison ith authentic racemic material, shown below: Chiralpak OD (4.6 x 250 mm), 98:2

hexanes:i-PrOH, 1.0 mL/min, λ = 254 nm.

(w), 1483 (w), 1451 (m), 1086 (s). H NMR (400 MHz, CDCl3, TMS): δ 7.32 (d, J = 7.0 Hz, 2H), 7.27−7.17 (m, 3H), 7.10-7.03 (m, 3H), 6.92 (d, J = 7.5 Hz, 1H), 6.66 (d, J = 15.8 Hz, 1H), 5.90 (dd, J = 15.8, 8.6 Hz, 1H), 5.83 (ddd, J = 17.4, 10.1, 8.4

Hz, 1H), 5.21 (dd, J = 17.4, 0.9 Hz, 1H), 5.07 (dd, J = 10.1, 1.3 Hz, 1H), 3.97 (d, J = 15.4 Hz, 1H), 3.78 (d, J = 15.4 Hz, 1H), 3.77−3.70 (m, 1H), 3.13 (ddd, J = 11.2, 8.6, 2.3 Hz, 1H), 3.06 (ddd, J = 11.2, 8.4, 2.6 Hz, 1H), 2.31 (s, 3H), 1.97−1.91 (m, 2H), 1.64 (br s, 1H), 1.60−1.48 (m, 2H). C NMR (100 MHz, CDCl3): δ 142.4, 140.0, 136.2, 135.6, 135.1, 130.1, 129.1, 128.1, 127.9, 127.3, 126.3, 126.1, 125.9, 115.9, 68.0, 64.3, 64.2, 55.1, 43.2, 43.0, 19.8. HRMS ES (m/z) mass Calcd for C23H28NO 334.2171 (M+H) , Found 334.2176. [α]D –106.2 (c 1.0, CHCl3) for a sample of >98% ee. T

NBn

OH

5.28

Me

ew

Chapter 5, Page 254

Authentic racemic 98% ee with Ru catalyst 5.2

Authentic racemic 98% ee with Ru catalyst 5.4

Method A for Ru-catalyzed

15.9 Hz, 1H), 5.83 (dd, J

Piperidine 5.29.111OH

AROM/CM was followed. IR (neat): 3355 (br), 2920 (m), 2839 (m), 1604 (m), 1511 (s), 1251 (m). 1H NMR (400 MHz, CDCl3, TMS): δ 7.28−7.14 (m, 5H), 7.05 (d, J = 8.8 Hz, 2H), 6.78 (d, J = 8.8 Hz, 2H), 6.38 (d, J =

= 15.9, 8.6 Hz, 1H), 5.81 (ddd, J = 17.4, 10.3, 8.4 Hz, 1H), 5.19 (dd, J = 17.4, 0.9 Hz, 1H), 5.05 (dd, J = 10.3, 1.5 Hz, 1H), 3.93 (d, J = 15.4 Hz, 1H), 3.78 (s, 3H), 3.75−3.67 (m, 1H), 3.73 (d, J = 15.4 Hz, 1H), 3.07 (ddd, J = 11.2, 8.6, 2.4 Hz, 1H), 3.03 (ddd, J = 11.2, 8.4, 2.6 Hz, 1H), 1.94−1.90 (m, 2H), 1.62 (br s, 1H), 1.57−1.46 (m, 2H). 13C NMR (100 MHz, CDCl3): δ 159.1, 142.5, 140.3, 132.1, 130.0, 129.7, 129.3, 127.9, 127.6, 126.3, 115.9, 114.0, 68.1, 64.5, 64.2, 55.4, 55.2, 43.2, 43.2. HRMS ES (m/z) mass Calcd for C23H28NO2 350.2120 (M+H)+, Found 350.2118. [α]D

25 –158.0 (c 1.0, CHCl3) for a sample of 98% ee. The enantiomeric purity of this compound was

(111) The 1H NMR spectra of this compound contains resi

OMe

NBn

5.29

dual water.

Chapter 5, Page 255

determined by HPLC analysis in comparison with authentic racemic material, shown below: Chiralpak OD (4.6 x 250 mm), 98:2 hexanes:i-PrOH, 1.0 mL/min, λ = 254 nm. Authentic racemic 98% ee with Ru catalyst 5.2

Authentic racemic 96% ee with Ru catalyst 5.4

for Ru-catalyzed asymmetric ringRepresentative procedure

metathesis/cross-metathesis r(

colorless oil (21.0 mg, 0.0540 mmol, 78%). IR (neat): 3358 (br), 2936 (w), 2848 (w), 1608 (w), 1325 (s), 1155 (m), 1124 (m). 1H NMR (400 MHz, CDCl3, TMS): δ 7.46 (d, J

-opening eactions, synthesis of piperidine 5.30: Ru complex 5.2 2.8 mg, 0.0035 mol) was dissolved in toluene (37.0 µL)

in a 4-mL vial. The resulting mixture was added by syringe to a solution of azabicycle 5.26 (15.0 mg, 0.0697 mmol), and p-trifluoromethylstyrene (51.0 µL, 0.350 mmol) in a 4-mL vial. The resulting solution was allowed

to stir for 12 h, at which time another portion of Ru complex 5.2 (2.8 mg, 0.0035 mmol) was added. The resulting mixture was allowed to stir for a further 12 h, upon which time the volatiles were removed in vacuo. The resulting dark brown residue was purified by silica gel column chromatography (99:1 CH2Cl2:MeOH) to furnish piperidine 5.30 as

NBn

CF35.30

OH

Chapter 5, Page 256

= 8.3 Hz, 2H), 7.28−7.16 (m, 5H), 7.10 (d, J = 8.3 Hz, 2H), 6.44 (d, J = 16.1 Hz, 1H), 5.99 (dd, J = 16.1, 8.8 Hz, 1H), 5.83 (ddd, J = 17.3, 10.2, 8.4 Hz, 1H), 5.21 (dd, J = 17.3,

3.75−3.70 (m, 1H), 3.62 (d, J = 15.4 Hz, 1H), 3.11 (ddd, J = 11.2, 8.8, 2.6 Hz, 1H), 3.04 (ddd, J = 11.2, 8.4, 2.6 Hz, 1H), 1.97−1.89 (m, 2H), 1.57 (br s, 1H), 1.59−1.49 (m, 2H). 13C NMR (100 MHz, CDCl3): δ 142.2, 140.6, 140.4, 137.1, 129.0, 128.6, 128.0, 126.5, 126.3, 125.4, 125.3, 116.2, 68.0, 64.6, 64.6, 55.7, 42.9 42.7, −5.1. HRMS ES (m/z) mass Calcd for C23H25F3NO 388.1888 (M+H)+, Found 388.1878. [α]D

25 –72.9 (c 0.5, CHCl3) for a sample of 97% ee. The enantiomeric purity of this compound was determined by HPLC analysis in comparison with authentic racemic material, shown below: Chiralpak OD (4.6 x 250 mm), 98:2 hexanes:i-PrOH, 1.0 mL/min, λ = 254 nm. Authentic racemic 97% ee with Ru catalyst 5.2

0.6 Hz, 1H), 5.08 (dd, J = 10.2, 1.3 Hz, 1H), 3.98 (d, J = 15.4 Hz, 1H),

Chapter 5, Page 257

Chapter 5, Page 258

Chapter 5, Page 259

Chapter 5, Page 260

Chapter 5, Page 261

Chapter 5, Page 262

Chapter 5, Page 263

Chapter 5, Page 264

Chapter 5, Page 265

Chapter 5, Page 266

Chapter 5, Page 267

Chapter 5, Page 268

Chapter 5, Page 269

Chapter 5, Page 270

Chapter 5, Page 271

Chapter 5, Page 272

Chapter 5, Page 273

Chapter 5, Page 274

Chapter 5, Page 275

Chapter 5, Page 276

Chapter 5, Page 277